Parkinson's Disease (PD) Detection Using Hand-Drawn Images

---

1 Project

• Motivation: In recent years, the prevalence of Parkinson's Disease (PD) has increased, highlighting the need for early and accurate detection. Hand-drawn patterns, are commonly used in clinical assessments to evaluate motor impairments associated with PD. By analyzing these images, we aim to leverage deep learning to develop an automated, non-invasive tool that complements clinical practices.

• Problem: Parkinson's Disease, a progressive neurodegenerative disorder, significantly impacts motor skills, making early detection vital for managing symptoms and improving quality of life. Can human judgment reliably detect subtle patterns in hand-drawn images? Could deep learning models provide an accurate and efficient solution for this binary classification task, distinguishing images as PD or non-PD?

• Approaches:

  • Deep Learning: Utilize CNN-based models, including MobileNetV2, VGG16, and DenseNet121, for feature extraction and image classification.

  • Traditional Models: Apply SVM, KNN, and Random Forest as baseline comparisons to evaluate the performance of deep learning models.

---

2 Data

• Acquisition: The dataset, "Parkinson's Disease Augmented Data of Handdrawings", is sourced from Kaggle. The data originates from the original dataset created by K Scott Mader, which contained 204 images. Through augmentation processes, including rotations (90°, 180°, 270°), vertical flipping (180°), and conversion to color images, 204 original images was increased to 3264 images (Anil Kumar, 2023).

• Description:

  • Folder: Dataset

  • Subfolders:

    • Healthy: 1,632 images
    • Parkinson: 1,632 images
      
      

  • Total Images: 3,264

  • Content: Hand-drawn spirals and waves.

  
  Reference
  
  Anil Kumar, B. (2023). Parkinson's Disease Augmented Data of Handdrawings [Dataset]. Kaggle. Retrieved from https://www.kaggle.com/datasets/banilkumar20phd7071/handwritten-parkinsons-disease-augmented-data/data.

---

Import Libraries

import os
import random
import warnings
warnings.filterwarnings("ignore")

import numpy as np
import pandas as pd
import tensorflow as tf
import matplotlib.pyplot as plt
import seaborn as sns
import cv2

SEED = 42
random.seed(SEED)
np.random.seed(SEED)
tf.random.set_seed(SEED)
os.environ['PYTHONHASHSEED'] = str(SEED)

Load Data

• Read image file paths from dataset folders.

• Assign corresponding labels: 0 for "Healthy", 1 for "Parkinson".

• Store image paths and labels in arrays.

# Define dataset path
dataset_path = '/kaggle/input/handwritten-parkinsons-disease-augmented-data/Dataset/Dataset'
output_dir = '/kaggle/working/preprocessed_dataset'

# Define categories and assign labels
classes = {'Healthy': 0, 'Parkinson': 1}

# Collect all image file paths and corresponding labels
image_paths = []
labels = []

for cls, label in classes.items():
    cls_path = os.path.join(dataset_path, cls)
    for img_name in os.listdir(cls_path):
        image_paths.append(os.path.join(cls_path, img_name))
        labels.append(label)

# Convert to NumPy arrays
image_paths = np.array(image_paths)
labels = np.array(labels)

print(f"Total images loaded: {len(image_paths)}")

Total images loaded: 3264

---

3 EDA

Step 1: Data Splitting (Train/Test)

• 80% for training and 20% for testing using train_test_split.

• Stratified split ensures a balanced ratio (Healthy vs. Parkinson) in both training and testing datasets.

from sklearn.model_selection import train_test_split

# Split dataset into training (80%) and testing (20%)
train_paths, test_paths, train_labels, test_labels = train_test_split(
    image_paths, labels, test_size=0.20, random_state=SEED, stratify=labels
)

print(f"Total images: {len(image_paths)} (Training: {len(train_paths)}, Testing: {len(test_paths)})")

Total images: 3264 (Training: 2611, Testing: 653)

# Visualization of label distribution
fig, axes = plt.subplots(1, 2, figsize=(13, 3.5), gridspec_kw={'wspace': 0.3})

# Define class labels
pie_labels = ['Healthy', 'Parkinson']

# Count occurrences of each label in the entire dataset
unique_labels, counts_labels = np.unique(labels, return_counts=True)

# Define colors
new_colors = ['lightblue', 'lightcoral']
edge_color = 'black'

# Prepare text annotations for Healthy and Parkinson counts
healthy_count = counts_labels[0]
parkinson_count = counts_labels[1]

# Plot pie chart for the entire dataset
wedges, texts, autotexts = axes[0].pie(
    counts_labels, labels=pie_labels, autopct='%1.1f%%', startangle=90, 
    colors=new_colors, wedgeprops={'edgecolor': edge_color, 'linewidth': 0.5}
)
axes[0].set_title('\nOverall Distribution\n')
axes[0].axis('equal')

# Annotate counts on pie chart
axes[0].text(-1.6, -1.15, f'Healthy: {healthy_count}', fontsize=8, color='black')
axes[0].text(-1.6, -1.3, f'Parkinson: {parkinson_count}', fontsize=8, color='black')

# Count occurrences in training and testing sets
train_healthy, train_parkinson = np.bincount(train_labels)
test_healthy, test_parkinson = np.bincount(test_labels)

# Plot bar chart for training and testing sets
bar_width = 0.3
x_positions = np.arange(len(['Training Set', 'Testing Set']))

bars_healthy = axes[1].bar(x_positions - bar_width / 2, [train_healthy, test_healthy], width=bar_width, 
                           label='Healthy', color=new_colors[0], edgecolor=edge_color, linewidth=0.5)
bars_parkinson = axes[1].bar(x_positions + bar_width / 2, [train_parkinson, test_parkinson], width=bar_width, 
                             label='Parkinson', color=new_colors[1], edgecolor=edge_color, linewidth=0.5)

# Annotate counts on bar chart
for bar in bars_healthy:
    axes[1].text(bar.get_x() + bar.get_width() / 2, bar.get_height() + 10, str(int(bar.get_height())), ha='center', fontsize=10)

for bar in bars_parkinson:
    axes[1].text(bar.get_x() + bar.get_width() / 2, bar.get_height() + 10, str(int(bar.get_height())), ha='center', fontsize=10)

axes[1].set_xlabel('Dataset')
axes[1].set_ylabel('Count')
axes[1].set_title('\nTraining vs Testing Distribution\n')
axes[1].set_xticks(x_positions)
axes[1].set_xticklabels(['Training Set', 'Testing Set'])
axes[1].grid(axis='y', linestyle='--', alpha=0.5)
axes[1].legend()

plt.tight_layout()
# plt.savefig('label_distribution.png', dpi=300)
plt.show()

The label distribution is perfectly balanced!

---

Step 2: Data Cleaning

• Check for Missing Data: Ensure that the image file exists and is accessible to avoid processing missing or unavailable files.

• Check Data Integrity: Detect and remove corrupted images that cannot be loaded to maintain data quality.

• Check for Extreme Values: Convert images to grayscale and analyze brightness levels to identify blank or overexposed images.

# Function to clean the dataset
def clean_image_data(image_paths):
    cleaned_paths = []
    for img_path in image_paths:
        # Check foe missing data
        if not os.path.exists(img_path):
            print(f"Missing file: {img_path}")
            continue

        image = cv2.imread(img_path)

        # Check data integrity
        if image is None:
            print(f"Corrupted image: {img_path}")
            continue
        
        # Check for extreme values
        gray_image = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)
        if np.mean(gray_image) < 10 or np.mean(gray_image) > 245:
            print(f"Potential noisy image: {img_path}")
            continue

        # Save cleaned image to a new directory
        cleaned_paths.append(img_path)

    return cleaned_paths

# Print the count before cleaning
print(f"Before cleaning - Total images: 3264")

# Apply the cleaning process to the dataset
train_paths = clean_image_data(train_paths)
test_paths = clean_image_data(test_paths)

# Print the final count after cleaning
print(f"After cleaning - Total images: {len(train_paths) + len(test_paths)}")

Before cleaning - Total images: 3264
After cleaning - Total images: 3264

The data cleaning process did not remove any images, showing that all images were valid and met the quality criteria perfectly!

---

Step 3: Data Preprocessing

1. Image Preprocessing and Visualizations:

• Resize all images to a fixed size of 256x256 pixels for consistency.

• Convert them into numerical arrays using img_to_array.

• Save them in train/ and test/ directories with subfolders for each class (0 for Healthy and 1 for Parkinson).

# Display some sample images
fig, axes = plt.subplots(2, 10, figsize=(15, 5))

# Get indices for each class
healthy_indices = np.where(labels == 0)[0][:10]
parkinson_indices = np.where(labels == 1)[0][:10]

for idx, img_idx in enumerate(healthy_indices):
    img_path = image_paths[img_idx]
    img = cv2.imread(img_path)
    img = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)

    axes[0, idx].imshow(img)
    axes[0, idx].set_title("Healthy")
    axes[0, idx].axis('off')

for idx, img_idx in enumerate(parkinson_indices):
    img_path = image_paths[img_idx]
    img = cv2.imread(img_path)
    img = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)

    axes[1, idx].imshow(img)
    axes[1, idx].set_title("Parkinson")
    axes[1, idx].axis('off')

plt.tight_layout()
# plt.savefig('images.png', dpi=300)
plt.show()

from tensorflow.keras.preprocessing.image import img_to_array

# Create output directories
if not os.path.exists(output_dir):
    os.makedirs(output_dir)
    os.makedirs(os.path.join(output_dir, 'train'))
    os.makedirs(os.path.join(output_dir, 'test'))

# Resizing and Saving
def preprocess_and_save(image_paths, labels, split_name):
    """Resize images, convert to arrays, and save to new directories."""
    target_size = (256, 256)
    for img_path, label in zip(image_paths, labels):
        image = cv2.imread(img_path)
        if image is not None:
            # Resize and convert image to array
            resized_image = cv2.resize(image, target_size)
            img_array = img_to_array(resized_image)

            # Define save path based on label
            class_dir = os.path.join(output_dir, split_name, str(label))
            if not os.path.exists(class_dir):
                os.makedirs(class_dir)

            # Save processed image
            img_name = os.path.basename(img_path)
            save_path = os.path.join(class_dir, img_name)
            cv2.imwrite(save_path, resized_image)

# Preprocess and save training and testing data
preprocess_and_save(train_paths, train_labels, 'train')
preprocess_and_save(test_paths, test_labels, 'test')

print("Data preprocessing completed. Resized images saved in train/test directories.")

Data preprocessing completed. Resized images saved in train/test directories.

# Define paths to preprocessed dataset
train_dir = os.path.join(output_dir, 'train')
test_dir = os.path.join(output_dir, 'test')

# Get processed Healthy and Parkinson images from the train set
healthy_images = [os.path.join(train_dir, '0', img) for img in os.listdir(os.path.join(train_dir, '0'))[:10]]
parkinson_images = [os.path.join(train_dir, '1', img) for img in os.listdir(os.path.join(train_dir, '1'))[:10]]

# Combine images and set titles
selected_images = healthy_images + parkinson_images
titles = ['Healthy'] * 10 + ['Parkinson'] * 10

# Display some processed images
fig, axes = plt.subplots(2, 10, figsize=(15, 5))

for idx, img_path in enumerate(selected_images):
    image = cv2.imread(img_path)
    image = cv2.cvtColor(image, cv2.COLOR_BGR2RGB)

    row = 0 if idx < 10 else 1
    col = idx % 10
    axes[row, col].imshow(image)
    axes[row, col].set_title(titles[idx])
    axes[row, col].axis('off')

plt.tight_layout()
# plt.savefig('processed_images.png', dpi=300)
plt.show()

All images have been resized to a consistent size. Well done!

---

• Edge Detection:

  • Healthy: The edges in the healthy samples appear smoother and more consistent, with well-defined and continuous lines forming clear patterns such as spirals and waves.

  • Parkinson: The edges in the Parkinson samples show irregularities, distortions, and jaggedness. There are noticeable tremors and inconsistencies in the drawn patterns, indicating potential motor control difficulties.

  • Conclusion: The edge detection results suggest that there are observable structural differences between healthy and Parkinson images, which could be useful features for classification purposes.

def edge_detection(img_path):
    image = cv2.imread(img_path, cv2.IMREAD_GRAYSCALE)
    edges = cv2.Canny(image, 50, 150)
    return edges

fig, axes = plt.subplots(2, 10, figsize=(15, 5))
for idx, img_path in enumerate(selected_images):
    edges = edge_detection(img_path)
    row = 0 if idx < 10 else 1
    col = idx % 10
    axes[row, col].imshow(edges, cmap='gray')
    axes[row, col].set_title(titles[idx])
    axes[row, col].axis('off')

plt.tight_layout()
# plt.savefig('edge_images.png', dpi=300)
plt.show()

---

• Pixel Intensity Distribution:

  • Healthy: Sharp peak near 250, indicating brighter pixels; minor peak around 150-200.

  • Parkinson: Similar trend but with a broader intensity range, especially in 150-200.

  • Conclusion: Parkinson images show more pixel variation, likely due to motor impairments affecting pen pressure and consistency.

# Initialize lists to store pixel values
healthy_pixels = []
parkinson_pixels = []

# Read all images from test set
healthy_dir = os.path.join(test_dir, '0')
parkinson_dir = os.path.join(test_dir, '1')

# Process Healthy images
for img_name in os.listdir(healthy_dir):
    img_path = os.path.join(healthy_dir, img_name)
    img = cv2.imread(img_path, cv2.IMREAD_GRAYSCALE)
    healthy_pixels.extend(img.ravel())  # Flatten and accumulate pixel values

# Process Parkinson images
for img_name in os.listdir(parkinson_dir):
    img_path = os.path.join(parkinson_dir, img_name)
    img = cv2.imread(img_path, cv2.IMREAD_GRAYSCALE)
    parkinson_pixels.extend(img.ravel())  # Flatten and accumulate pixel values

# Plot KDEs
plt.figure(figsize=(8, 5))

# Plot Healthy pixels
sns.kdeplot(healthy_pixels, color='skyblue', fill=True, alpha=0.5, label='Healthy')

# Plot Parkinson pixels
sns.kdeplot(parkinson_pixels, color='salmon', fill=True, alpha=0.5, label='Parkinson')

plt.title('\nPixel Intensity Distribution\n')
plt.xlabel('Pixel Intensity')
plt.ylabel('Density')
plt.legend(loc='upper left', bbox_to_anchor=(0, 1))

plt.savefig('pixel_distribution.png', dpi=300)
plt.show()

---

2. Image Preparation (for Deep Learning):

• Data Normalization:

  • Apply pixel value scaling to the range [0,1] using rescale=1./255 to normalize image data.
    
    

• Load Training data:

  • Resize images to 224x224 pixels to match input requirements for pretrained models
  • Use a batch size of 16 for efficient processing.
  • Set class_mode='binary' to handle two classes (0: Healthy, 1: Parkinson).
    
    

• Load Testing data:

  • Load images from the testing directory with the same preprocessing steps as training.
  • Ensure consistent target size and batch size for evaluation purposes.
  • Keep image ordering consistent by setting shuffle=False.

from tensorflow.keras.preprocessing.image import ImageDataGenerator

# Data normalization
datagen = ImageDataGenerator(rescale=1./255)

# Load training data
train_generator = datagen.flow_from_directory(
    train_dir,
    target_size=(224, 224), # Adjusted for pretrained models
    batch_size=16,
    class_mode='binary',  # Binarized labels (0 and 1)
    shuffle=True,  # Shuffle training data
    seed=SEED
)

# Load test data
test_generator = datagen.flow_from_directory(
    test_dir,
    target_size=(224, 224),
    batch_size=16,
    class_mode='binary',
    shuffle=False,  # Keep evaluation order consistent
    seed=SEED
) 

# Check class indices
print("Class indices:", train_generator.class_indices)

Found 2611 images belonging to 2 classes.
Found 653 images belonging to 2 classes.
Class indices: {'0': 0, '1': 1}

---

3. Feature Extraction (for Traditional Models):

• Image Preprocessing:

  • Load and resize images to 224x224 to match the VGG16 input size.
  • Convert images to numerical arrays using image.img_to_array().
  • Using preprocess_input() to normalize pixel values.
    
    

• Feature Extraction:

  • Use the VGG16 model to extract meaningful features from the images.
  • Flatten the extracted features into a 1D feature vector.
    
    

• Labeling: Assign class labels (0 for Healthy, 1 for Parkinson).

• Storage: Save extracted features and labels as NumPy arrays for ML models.

from tensorflow.keras.applications import VGG16
from tensorflow.keras.preprocessing import image
from tensorflow.keras.applications.vgg16 import preprocess_input

# Load pre-trained VGG16 model without the top layers (only feature extractor)
base_model = VGG16(weights='imagenet', include_top=False, input_shape=(224, 224, 3))

def extract_features(image_dir):
    features = []
    labels = []

    for label in ['0', '1']:  # 0: Healthy, 1: Parkinson
        class_dir = os.path.join(image_dir, label)
        for img_name in os.listdir(class_dir):
            img_path = os.path.join(class_dir, img_name)
            img = image.load_img(img_path, target_size=(224, 224))
            img_array = image.img_to_array(img)
            img_array = preprocess_input(np.expand_dims(img_array, axis=0))

            # Extract features from VGG16
            feature = base_model.predict(img_array).flatten()
            features.append(feature)
            labels.append(int(label))
            
    return np.array(features), np.array(labels)

# Extract features
X_train, y_train = extract_features(os.path.join(output_dir, 'train'))
X_test, y_test = extract_features(os.path.join(output_dir, 'test'))

Downloading data from https://storage.googleapis.com/tensorflow/keras-applications/vgg16/vgg16_weights_tf_dim_ordering_tf_kernels_notop.h5
58889256/58889256 ━━━━━━━━━━━━━━━━━━━━ 0s 0us/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 3s 3s/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 22ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 17ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 17ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 17ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 17ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 17ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 17ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 17ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 24ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 25ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 32ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 23ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 22ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 17ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 22ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 17ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 17ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 29ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 22ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 17ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 17ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 22ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 24ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 17ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 17ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 17ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 17ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 22ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 26ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 17ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 22ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 17ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 25ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 22ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 26ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 22ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 31ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 23ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 17ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 17ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 26ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 17ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 17ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 27ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 22ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 22ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 23ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 24ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 22ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 22ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 22ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 22ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 22ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 22ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 26ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 22ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 25ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 23ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 28ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 24ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 25ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 22ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 22ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 18ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 21ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 19ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 20ms/step

# Check the shape of the feature matrices
print("Feature shape:", X_train.shape, X_test.shape)

Feature shape: (2611, 25088) (653, 25088)

---

4 Models

Step 1: Deep Learning

Overview:

1. CNN (Custom): A basic custom CNN is used to build a model from scratch, providing a foundational approach to extract features from hand-drawn images and classify them. This allows full control over the architecture and serves as a baseline to compare the performance of more advanced pre-trained models.

2. MobileNetV2 (Transfer Learning): MobileNetV2 is a pre-trained model selected for its lightweight and efficient design. Using transfer learning, it can leverage features learned from large-scale datasets, such as ImageNet, to adapt to the Parkinson's disease dataset. This is particularly useful for scenarios with limited data, ensuring better performance with reduced training time and resources.

3. VGG16 (Transfer Learning): VGG16 is chosen as a pre-trained model for its deep and consistent architecture, which has been widely validated in image classification tasks. Transfer learning enables the model to utilize its pre-learned feature representations, making it effective for detecting patterns in the hand-drawn images with minimal tuning.

4. DenseNet121 (Transfer Learning): DenseNet121 is included for its dense connectivity, which promotes feature reuse and enhances gradient flow. As a pre-trained model, it leverages its robust feature extraction capabilities to identify subtle patterns in the dataset, making it a strong candidate for complex classification tasks.

---

1. CNN:

• Input Layer: Accepts images of shape (224, 224, 3).

• Convolutional-Pooling Blocks:

  • Block 1: 32 filters (3x3), ReLU activation, MaxPooling (2x2).
  • Block 2: 64 filters (3x3), ReLU activation, MaxPooling (2x2).
  • Block 3: 128 filters (3x3), ReLU activation, MaxPooling (2x2).
    
    

• Flatten Layer: Converts 2D features to 1D.

• Fully Connected Layers:

  • Dense (256 neurons, ReLU activation).
  • Dropout (0.5 to prevent overfitting).
    
    

• Output Layer: Dense (1 neuron, sigmoid activation) for binary classification.

from tensorflow.keras.models import Sequential
from tensorflow.keras.layers import Conv2D, MaxPooling2D, Flatten, Dense, Dropout

# Build the CNN model
model = Sequential([
    Conv2D(32, (3,3), activation='relu', input_shape=(224, 224, 3)),
    MaxPooling2D(2,2),

    Conv2D(64, (3,3), activation='relu'),
    MaxPooling2D(2,2),

    Conv2D(128, (3,3), activation='relu'),
    MaxPooling2D(2,2),

    Flatten(),
    Dense(256, activation='relu'),
    Dropout(0.5),
    Dense(1, activation='sigmoid')
])

# Compile the model
model.compile(optimizer='adam', loss='binary_crossentropy', metrics=['accuracy'])

# Model summary
model.summary()

Model: "sequential_2"

┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━┓
┃ Layer (type)                         ┃ Output Shape                ┃         Param # ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━┩
│ conv2d_6 (Conv2D)                    │ (None, 222, 222, 32)        │             896 │
├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤
│ max_pooling2d_6 (MaxPooling2D)       │ (None, 111, 111, 32)        │               0 │
├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤
│ conv2d_7 (Conv2D)                    │ (None, 109, 109, 64)        │          18,496 │
├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤
│ max_pooling2d_7 (MaxPooling2D)       │ (None, 54, 54, 64)          │               0 │
├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤
│ conv2d_8 (Conv2D)                    │ (None, 52, 52, 128)         │          73,856 │
├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤
│ max_pooling2d_8 (MaxPooling2D)       │ (None, 26, 26, 128)         │               0 │
├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤
│ flatten_2 (Flatten)                  │ (None, 86528)               │               0 │
├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤
│ dense_12 (Dense)                     │ (None, 256)                 │      22,151,424 │
├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤
│ dropout_6 (Dropout)                  │ (None, 256)                 │               0 │
├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤
│ dense_13 (Dense)                     │ (None, 1)                   │             257 │
└──────────────────────────────────────┴─────────────────────────────┴─────────────────┘

 Total params: 22,244,929 (84.86 MB)

 Trainable params: 22,244,929 (84.86 MB)

 Non-trainable params: 0 (0.00 B)

import time
from tensorflow.keras.callbacks import EarlyStopping

# Early stopping
early_stopping = EarlyStopping(monitor='val_loss', patience=5, restore_best_weights=True)

# Start timer
start_time = time.time()

# Train the model
history = model.fit(
    train_generator,
    epochs=20,
    validation_data=test_generator,
    callbacks=[early_stopping]
)

# End timer
end_time = time.time()

# Calculate training duration
training_time = end_time - start_time
minutes, seconds = divmod(training_time, 60)
print("-" * 120)
print(f"Training completed in {int(minutes)} minutes and {int(seconds)} seconds.")

# Plot accuracy and loss graphs
plt.figure(figsize=(12, 3))

# Plot Accuracy
plt.subplot(1, 2, 1)
plt.plot(history.history['accuracy'], label='Train Accuracy')
plt.plot(history.history['val_accuracy'], label='Test Accuracy')
plt.xlabel('Epochs')
plt.ylabel('Accuracy')
plt.legend()
plt.title('\nTraining vs Test Accuracy\n')

# Plot Loss
plt.subplot(1, 2, 2)
plt.plot(history.history['loss'], label='Train Loss')
plt.plot(history.history['val_loss'], label='Test Loss')
plt.xlabel('Epochs')
plt.ylabel('Loss')
plt.legend()
plt.title('\nTraining vs Test Loss\n')

plt.show()

Epoch 1/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 16s 72ms/step - accuracy: 0.5166 - loss: 0.7416 - val_accuracy: 0.5161 - val_loss: 0.6864
Epoch 2/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 10s 60ms/step - accuracy: 0.5812 - loss: 0.6790 - val_accuracy: 0.7213 - val_loss: 0.5922
Epoch 3/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 10s 60ms/step - accuracy: 0.7269 - loss: 0.5341 - val_accuracy: 0.7580 - val_loss: 0.5210
Epoch 4/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 10s 60ms/step - accuracy: 0.8357 - loss: 0.3792 - val_accuracy: 0.8331 - val_loss: 0.3755
Epoch 5/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 10s 59ms/step - accuracy: 0.9192 - loss: 0.2152 - val_accuracy: 0.8882 - val_loss: 0.3278
Epoch 6/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 10s 61ms/step - accuracy: 0.9537 - loss: 0.1449 - val_accuracy: 0.9234 - val_loss: 0.2875
Epoch 7/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 10s 59ms/step - accuracy: 0.9767 - loss: 0.0700 - val_accuracy: 0.9372 - val_loss: 0.3206
Epoch 8/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 10s 60ms/step - accuracy: 0.9800 - loss: 0.0655 - val_accuracy: 0.9433 - val_loss: 0.2483
Epoch 9/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 10s 58ms/step - accuracy: 0.9758 - loss: 0.0816 - val_accuracy: 0.9525 - val_loss: 0.2692
Epoch 10/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 10s 60ms/step - accuracy: 0.9930 - loss: 0.0279 - val_accuracy: 0.9571 - val_loss: 0.2627
Epoch 11/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 11s 62ms/step - accuracy: 0.9869 - loss: 0.0416 - val_accuracy: 0.9571 - val_loss: 0.2408
Epoch 12/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 10s 60ms/step - accuracy: 0.9935 - loss: 0.0177 - val_accuracy: 0.9632 - val_loss: 0.2225
Epoch 13/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 10s 58ms/step - accuracy: 0.9957 - loss: 0.0229 - val_accuracy: 0.9587 - val_loss: 0.3222
Epoch 14/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 10s 59ms/step - accuracy: 0.9962 - loss: 0.0129 - val_accuracy: 0.9587 - val_loss: 0.2767
Epoch 15/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 10s 59ms/step - accuracy: 0.9954 - loss: 0.0118 - val_accuracy: 0.9556 - val_loss: 0.3930
Epoch 16/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 10s 61ms/step - accuracy: 0.9951 - loss: 0.0270 - val_accuracy: 0.9602 - val_loss: 0.3254
Epoch 17/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 10s 60ms/step - accuracy: 0.9969 - loss: 0.0116 - val_accuracy: 0.9632 - val_loss: 0.3894
------------------------------------------------------------------------------------------------------------------------
Training completed in 2 minutes and 59 seconds.

from sklearn.metrics import confusion_matrix, roc_auc_score, accuracy_score, f1_score

# Predict the test set
y_pred_prob = model.predict(test_generator)
y_pred = (y_pred_prob > 0.5).astype(int)

# Get true labels
y_true = test_generator.classes

# Confusion Matrix
conf_matrix = confusion_matrix(y_true, y_pred)
plt.figure(figsize=(4, 3))
sns.heatmap(conf_matrix, annot=True, fmt='d', cmap='Blues', 
            xticklabels=['Healthy', 'Parkinson'], 
            yticklabels=['Healthy', 'Parkinson'], 
            cbar=False)
plt.xlabel('Predicted')
plt.ylabel('Actual')
plt.title('\nTest Data\n')
plt.show()

# Evaluation Metrics
accuracy = accuracy_score(y_true, y_pred)
f1 = f1_score(y_true, y_pred)
auc = roc_auc_score(y_true, y_pred_prob)

print(f'Accuracy: {accuracy:.2f}')
print(f'F1 Score: {f1:.2f}')
print(f'AUC Score: {auc:.2f}')

41/41 ━━━━━━━━━━━━━━━━━━━━ 2s 50ms/step

Accuracy: 0.96
F1 Score: 0.96
AUC Score: 0.98

---

2. MobileNetV2:

• Base Model:

  • Uses MobileNetV2 pre-trained on ImageNet as the feature extractor.
  • The top classification layers are excluded (include_top=False).
  • All layers are frozen to retain pre-trained weights (base_model.trainable = False).
    
    

• Custom Classification Head:

  • GlobalAveragePooling2D: Reduces feature maps to a single value.
  • Dense (256 neurons, ReLU activation).
  • Dropout (0.5 to prevent overfitting).
  • Dense (1 neuron, sigmoid activation) for binary classification.

from tensorflow.keras.applications import MobileNetV2
from tensorflow.keras.layers import GlobalAveragePooling2D
from tensorflow.keras.models import Model

# Load MobileNetV2 as the base model (excluding the top classification layers)
base_model = MobileNetV2(weights='imagenet', include_top=False, input_shape=(224, 224, 3))

# Freeze all layers of the pre-trained model
base_model.trainable = False

# Add new classification layers on top of MobileNetV2
x = base_model.output
x = GlobalAveragePooling2D()(x)
x = Dense(256, activation='relu')(x)
x = Dropout(0.5)(x)
output = Dense(1, activation='sigmoid')(x)

# Build the complete model
mobilenet_model = Model(inputs=base_model.input, outputs=output)

# Compile the model
mobilenet_model.compile(optimizer='adam', loss='binary_crossentropy', metrics=['accuracy'])

# Model summary
mobilenet_model.summary()

Model: "functional_5"

┏━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━┓
┃ Layer (type)              ┃ Output Shape           ┃        Param # ┃ Connected to           ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━┩
│ input_layer_6             │ (None, 224, 224, 3)    │              0 │ -                      │
│ (InputLayer)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ Conv1 (Conv2D)            │ (None, 112, 112, 32)   │            864 │ input_layer_6[0][0]    │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ bn_Conv1                  │ (None, 112, 112, 32)   │            128 │ Conv1[0][0]            │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ Conv1_relu (ReLU)         │ (None, 112, 112, 32)   │              0 │ bn_Conv1[0][0]         │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ expanded_conv_depthwise   │ (None, 112, 112, 32)   │            288 │ Conv1_relu[0][0]       │
│ (DepthwiseConv2D)         │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ expanded_conv_depthwise_… │ (None, 112, 112, 32)   │            128 │ expanded_conv_depthwi… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ expanded_conv_depthwise_… │ (None, 112, 112, 32)   │              0 │ expanded_conv_depthwi… │
│ (ReLU)                    │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ expanded_conv_project     │ (None, 112, 112, 16)   │            512 │ expanded_conv_depthwi… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ expanded_conv_project_BN  │ (None, 112, 112, 16)   │             64 │ expanded_conv_project… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_1_expand (Conv2D)   │ (None, 112, 112, 96)   │          1,536 │ expanded_conv_project… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_1_expand_BN         │ (None, 112, 112, 96)   │            384 │ block_1_expand[0][0]   │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_1_expand_relu       │ (None, 112, 112, 96)   │              0 │ block_1_expand_BN[0][… │
│ (ReLU)                    │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_1_pad               │ (None, 113, 113, 96)   │              0 │ block_1_expand_relu[0… │
│ (ZeroPadding2D)           │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_1_depthwise         │ (None, 56, 56, 96)     │            864 │ block_1_pad[0][0]      │
│ (DepthwiseConv2D)         │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_1_depthwise_BN      │ (None, 56, 56, 96)     │            384 │ block_1_depthwise[0][… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_1_depthwise_relu    │ (None, 56, 56, 96)     │              0 │ block_1_depthwise_BN[… │
│ (ReLU)                    │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_1_project (Conv2D)  │ (None, 56, 56, 24)     │          2,304 │ block_1_depthwise_rel… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_1_project_BN        │ (None, 56, 56, 24)     │             96 │ block_1_project[0][0]  │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_2_expand (Conv2D)   │ (None, 56, 56, 144)    │          3,456 │ block_1_project_BN[0]… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_2_expand_BN         │ (None, 56, 56, 144)    │            576 │ block_2_expand[0][0]   │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_2_expand_relu       │ (None, 56, 56, 144)    │              0 │ block_2_expand_BN[0][… │
│ (ReLU)                    │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_2_depthwise         │ (None, 56, 56, 144)    │          1,296 │ block_2_expand_relu[0… │
│ (DepthwiseConv2D)         │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_2_depthwise_BN      │ (None, 56, 56, 144)    │            576 │ block_2_depthwise[0][… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_2_depthwise_relu    │ (None, 56, 56, 144)    │              0 │ block_2_depthwise_BN[… │
│ (ReLU)                    │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_2_project (Conv2D)  │ (None, 56, 56, 24)     │          3,456 │ block_2_depthwise_rel… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_2_project_BN        │ (None, 56, 56, 24)     │             96 │ block_2_project[0][0]  │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_2_add (Add)         │ (None, 56, 56, 24)     │              0 │ block_1_project_BN[0]… │
│                           │                        │                │ block_2_project_BN[0]… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_3_expand (Conv2D)   │ (None, 56, 56, 144)    │          3,456 │ block_2_add[0][0]      │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_3_expand_BN         │ (None, 56, 56, 144)    │            576 │ block_3_expand[0][0]   │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_3_expand_relu       │ (None, 56, 56, 144)    │              0 │ block_3_expand_BN[0][… │
│ (ReLU)                    │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_3_pad               │ (None, 57, 57, 144)    │              0 │ block_3_expand_relu[0… │
│ (ZeroPadding2D)           │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_3_depthwise         │ (None, 28, 28, 144)    │          1,296 │ block_3_pad[0][0]      │
│ (DepthwiseConv2D)         │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_3_depthwise_BN      │ (None, 28, 28, 144)    │            576 │ block_3_depthwise[0][… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_3_depthwise_relu    │ (None, 28, 28, 144)    │              0 │ block_3_depthwise_BN[… │
│ (ReLU)                    │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_3_project (Conv2D)  │ (None, 28, 28, 32)     │          4,608 │ block_3_depthwise_rel… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_3_project_BN        │ (None, 28, 28, 32)     │            128 │ block_3_project[0][0]  │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_4_expand (Conv2D)   │ (None, 28, 28, 192)    │          6,144 │ block_3_project_BN[0]… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_4_expand_BN         │ (None, 28, 28, 192)    │            768 │ block_4_expand[0][0]   │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_4_expand_relu       │ (None, 28, 28, 192)    │              0 │ block_4_expand_BN[0][… │
│ (ReLU)                    │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_4_depthwise         │ (None, 28, 28, 192)    │          1,728 │ block_4_expand_relu[0… │
│ (DepthwiseConv2D)         │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_4_depthwise_BN      │ (None, 28, 28, 192)    │            768 │ block_4_depthwise[0][… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_4_depthwise_relu    │ (None, 28, 28, 192)    │              0 │ block_4_depthwise_BN[… │
│ (ReLU)                    │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_4_project (Conv2D)  │ (None, 28, 28, 32)     │          6,144 │ block_4_depthwise_rel… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_4_project_BN        │ (None, 28, 28, 32)     │            128 │ block_4_project[0][0]  │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_4_add (Add)         │ (None, 28, 28, 32)     │              0 │ block_3_project_BN[0]… │
│                           │                        │                │ block_4_project_BN[0]… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_5_expand (Conv2D)   │ (None, 28, 28, 192)    │          6,144 │ block_4_add[0][0]      │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_5_expand_BN         │ (None, 28, 28, 192)    │            768 │ block_5_expand[0][0]   │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_5_expand_relu       │ (None, 28, 28, 192)    │              0 │ block_5_expand_BN[0][… │
│ (ReLU)                    │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_5_depthwise         │ (None, 28, 28, 192)    │          1,728 │ block_5_expand_relu[0… │
│ (DepthwiseConv2D)         │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_5_depthwise_BN      │ (None, 28, 28, 192)    │            768 │ block_5_depthwise[0][… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_5_depthwise_relu    │ (None, 28, 28, 192)    │              0 │ block_5_depthwise_BN[… │
│ (ReLU)                    │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_5_project (Conv2D)  │ (None, 28, 28, 32)     │          6,144 │ block_5_depthwise_rel… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_5_project_BN        │ (None, 28, 28, 32)     │            128 │ block_5_project[0][0]  │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_5_add (Add)         │ (None, 28, 28, 32)     │              0 │ block_4_add[0][0],     │
│                           │                        │                │ block_5_project_BN[0]… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_6_expand (Conv2D)   │ (None, 28, 28, 192)    │          6,144 │ block_5_add[0][0]      │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_6_expand_BN         │ (None, 28, 28, 192)    │            768 │ block_6_expand[0][0]   │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_6_expand_relu       │ (None, 28, 28, 192)    │              0 │ block_6_expand_BN[0][… │
│ (ReLU)                    │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_6_pad               │ (None, 29, 29, 192)    │              0 │ block_6_expand_relu[0… │
│ (ZeroPadding2D)           │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_6_depthwise         │ (None, 14, 14, 192)    │          1,728 │ block_6_pad[0][0]      │
│ (DepthwiseConv2D)         │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_6_depthwise_BN      │ (None, 14, 14, 192)    │            768 │ block_6_depthwise[0][… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_6_depthwise_relu    │ (None, 14, 14, 192)    │              0 │ block_6_depthwise_BN[… │
│ (ReLU)                    │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_6_project (Conv2D)  │ (None, 14, 14, 64)     │         12,288 │ block_6_depthwise_rel… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_6_project_BN        │ (None, 14, 14, 64)     │            256 │ block_6_project[0][0]  │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_7_expand (Conv2D)   │ (None, 14, 14, 384)    │         24,576 │ block_6_project_BN[0]… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_7_expand_BN         │ (None, 14, 14, 384)    │          1,536 │ block_7_expand[0][0]   │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_7_expand_relu       │ (None, 14, 14, 384)    │              0 │ block_7_expand_BN[0][… │
│ (ReLU)                    │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_7_depthwise         │ (None, 14, 14, 384)    │          3,456 │ block_7_expand_relu[0… │
│ (DepthwiseConv2D)         │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_7_depthwise_BN      │ (None, 14, 14, 384)    │          1,536 │ block_7_depthwise[0][… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_7_depthwise_relu    │ (None, 14, 14, 384)    │              0 │ block_7_depthwise_BN[… │
│ (ReLU)                    │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_7_project (Conv2D)  │ (None, 14, 14, 64)     │         24,576 │ block_7_depthwise_rel… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_7_project_BN        │ (None, 14, 14, 64)     │            256 │ block_7_project[0][0]  │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_7_add (Add)         │ (None, 14, 14, 64)     │              0 │ block_6_project_BN[0]… │
│                           │                        │                │ block_7_project_BN[0]… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_8_expand (Conv2D)   │ (None, 14, 14, 384)    │         24,576 │ block_7_add[0][0]      │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_8_expand_BN         │ (None, 14, 14, 384)    │          1,536 │ block_8_expand[0][0]   │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_8_expand_relu       │ (None, 14, 14, 384)    │              0 │ block_8_expand_BN[0][… │
│ (ReLU)                    │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_8_depthwise         │ (None, 14, 14, 384)    │          3,456 │ block_8_expand_relu[0… │
│ (DepthwiseConv2D)         │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_8_depthwise_BN      │ (None, 14, 14, 384)    │          1,536 │ block_8_depthwise[0][… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_8_depthwise_relu    │ (None, 14, 14, 384)    │              0 │ block_8_depthwise_BN[… │
│ (ReLU)                    │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_8_project (Conv2D)  │ (None, 14, 14, 64)     │         24,576 │ block_8_depthwise_rel… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_8_project_BN        │ (None, 14, 14, 64)     │            256 │ block_8_project[0][0]  │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_8_add (Add)         │ (None, 14, 14, 64)     │              0 │ block_7_add[0][0],     │
│                           │                        │                │ block_8_project_BN[0]… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_9_expand (Conv2D)   │ (None, 14, 14, 384)    │         24,576 │ block_8_add[0][0]      │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_9_expand_BN         │ (None, 14, 14, 384)    │          1,536 │ block_9_expand[0][0]   │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_9_expand_relu       │ (None, 14, 14, 384)    │              0 │ block_9_expand_BN[0][… │
│ (ReLU)                    │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_9_depthwise         │ (None, 14, 14, 384)    │          3,456 │ block_9_expand_relu[0… │
│ (DepthwiseConv2D)         │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_9_depthwise_BN      │ (None, 14, 14, 384)    │          1,536 │ block_9_depthwise[0][… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_9_depthwise_relu    │ (None, 14, 14, 384)    │              0 │ block_9_depthwise_BN[… │
│ (ReLU)                    │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_9_project (Conv2D)  │ (None, 14, 14, 64)     │         24,576 │ block_9_depthwise_rel… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_9_project_BN        │ (None, 14, 14, 64)     │            256 │ block_9_project[0][0]  │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_9_add (Add)         │ (None, 14, 14, 64)     │              0 │ block_8_add[0][0],     │
│                           │                        │                │ block_9_project_BN[0]… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_10_expand (Conv2D)  │ (None, 14, 14, 384)    │         24,576 │ block_9_add[0][0]      │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_10_expand_BN        │ (None, 14, 14, 384)    │          1,536 │ block_10_expand[0][0]  │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_10_expand_relu      │ (None, 14, 14, 384)    │              0 │ block_10_expand_BN[0]… │
│ (ReLU)                    │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_10_depthwise        │ (None, 14, 14, 384)    │          3,456 │ block_10_expand_relu[… │
│ (DepthwiseConv2D)         │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_10_depthwise_BN     │ (None, 14, 14, 384)    │          1,536 │ block_10_depthwise[0]… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_10_depthwise_relu   │ (None, 14, 14, 384)    │              0 │ block_10_depthwise_BN… │
│ (ReLU)                    │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_10_project (Conv2D) │ (None, 14, 14, 96)     │         36,864 │ block_10_depthwise_re… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_10_project_BN       │ (None, 14, 14, 96)     │            384 │ block_10_project[0][0] │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_11_expand (Conv2D)  │ (None, 14, 14, 576)    │         55,296 │ block_10_project_BN[0… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_11_expand_BN        │ (None, 14, 14, 576)    │          2,304 │ block_11_expand[0][0]  │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_11_expand_relu      │ (None, 14, 14, 576)    │              0 │ block_11_expand_BN[0]… │
│ (ReLU)                    │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_11_depthwise        │ (None, 14, 14, 576)    │          5,184 │ block_11_expand_relu[… │
│ (DepthwiseConv2D)         │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_11_depthwise_BN     │ (None, 14, 14, 576)    │          2,304 │ block_11_depthwise[0]… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_11_depthwise_relu   │ (None, 14, 14, 576)    │              0 │ block_11_depthwise_BN… │
│ (ReLU)                    │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_11_project (Conv2D) │ (None, 14, 14, 96)     │         55,296 │ block_11_depthwise_re… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_11_project_BN       │ (None, 14, 14, 96)     │            384 │ block_11_project[0][0] │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_11_add (Add)        │ (None, 14, 14, 96)     │              0 │ block_10_project_BN[0… │
│                           │                        │                │ block_11_project_BN[0… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_12_expand (Conv2D)  │ (None, 14, 14, 576)    │         55,296 │ block_11_add[0][0]     │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_12_expand_BN        │ (None, 14, 14, 576)    │          2,304 │ block_12_expand[0][0]  │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_12_expand_relu      │ (None, 14, 14, 576)    │              0 │ block_12_expand_BN[0]… │
│ (ReLU)                    │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_12_depthwise        │ (None, 14, 14, 576)    │          5,184 │ block_12_expand_relu[… │
│ (DepthwiseConv2D)         │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_12_depthwise_BN     │ (None, 14, 14, 576)    │          2,304 │ block_12_depthwise[0]… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_12_depthwise_relu   │ (None, 14, 14, 576)    │              0 │ block_12_depthwise_BN… │
│ (ReLU)                    │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_12_project (Conv2D) │ (None, 14, 14, 96)     │         55,296 │ block_12_depthwise_re… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_12_project_BN       │ (None, 14, 14, 96)     │            384 │ block_12_project[0][0] │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_12_add (Add)        │ (None, 14, 14, 96)     │              0 │ block_11_add[0][0],    │
│                           │                        │                │ block_12_project_BN[0… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_13_expand (Conv2D)  │ (None, 14, 14, 576)    │         55,296 │ block_12_add[0][0]     │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_13_expand_BN        │ (None, 14, 14, 576)    │          2,304 │ block_13_expand[0][0]  │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_13_expand_relu      │ (None, 14, 14, 576)    │              0 │ block_13_expand_BN[0]… │
│ (ReLU)                    │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_13_pad              │ (None, 15, 15, 576)    │              0 │ block_13_expand_relu[… │
│ (ZeroPadding2D)           │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_13_depthwise        │ (None, 7, 7, 576)      │          5,184 │ block_13_pad[0][0]     │
│ (DepthwiseConv2D)         │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_13_depthwise_BN     │ (None, 7, 7, 576)      │          2,304 │ block_13_depthwise[0]… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_13_depthwise_relu   │ (None, 7, 7, 576)      │              0 │ block_13_depthwise_BN… │
│ (ReLU)                    │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_13_project (Conv2D) │ (None, 7, 7, 160)      │         92,160 │ block_13_depthwise_re… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_13_project_BN       │ (None, 7, 7, 160)      │            640 │ block_13_project[0][0] │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_14_expand (Conv2D)  │ (None, 7, 7, 960)      │        153,600 │ block_13_project_BN[0… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_14_expand_BN        │ (None, 7, 7, 960)      │          3,840 │ block_14_expand[0][0]  │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_14_expand_relu      │ (None, 7, 7, 960)      │              0 │ block_14_expand_BN[0]… │
│ (ReLU)                    │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_14_depthwise        │ (None, 7, 7, 960)      │          8,640 │ block_14_expand_relu[… │
│ (DepthwiseConv2D)         │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_14_depthwise_BN     │ (None, 7, 7, 960)      │          3,840 │ block_14_depthwise[0]… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_14_depthwise_relu   │ (None, 7, 7, 960)      │              0 │ block_14_depthwise_BN… │
│ (ReLU)                    │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_14_project (Conv2D) │ (None, 7, 7, 160)      │        153,600 │ block_14_depthwise_re… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_14_project_BN       │ (None, 7, 7, 160)      │            640 │ block_14_project[0][0] │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_14_add (Add)        │ (None, 7, 7, 160)      │              0 │ block_13_project_BN[0… │
│                           │                        │                │ block_14_project_BN[0… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_15_expand (Conv2D)  │ (None, 7, 7, 960)      │        153,600 │ block_14_add[0][0]     │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_15_expand_BN        │ (None, 7, 7, 960)      │          3,840 │ block_15_expand[0][0]  │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_15_expand_relu      │ (None, 7, 7, 960)      │              0 │ block_15_expand_BN[0]… │
│ (ReLU)                    │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_15_depthwise        │ (None, 7, 7, 960)      │          8,640 │ block_15_expand_relu[… │
│ (DepthwiseConv2D)         │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_15_depthwise_BN     │ (None, 7, 7, 960)      │          3,840 │ block_15_depthwise[0]… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_15_depthwise_relu   │ (None, 7, 7, 960)      │              0 │ block_15_depthwise_BN… │
│ (ReLU)                    │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_15_project (Conv2D) │ (None, 7, 7, 160)      │        153,600 │ block_15_depthwise_re… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_15_project_BN       │ (None, 7, 7, 160)      │            640 │ block_15_project[0][0] │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_15_add (Add)        │ (None, 7, 7, 160)      │              0 │ block_14_add[0][0],    │
│                           │                        │                │ block_15_project_BN[0… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_16_expand (Conv2D)  │ (None, 7, 7, 960)      │        153,600 │ block_15_add[0][0]     │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_16_expand_BN        │ (None, 7, 7, 960)      │          3,840 │ block_16_expand[0][0]  │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_16_expand_relu      │ (None, 7, 7, 960)      │              0 │ block_16_expand_BN[0]… │
│ (ReLU)                    │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_16_depthwise        │ (None, 7, 7, 960)      │          8,640 │ block_16_expand_relu[… │
│ (DepthwiseConv2D)         │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_16_depthwise_BN     │ (None, 7, 7, 960)      │          3,840 │ block_16_depthwise[0]… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_16_depthwise_relu   │ (None, 7, 7, 960)      │              0 │ block_16_depthwise_BN… │
│ (ReLU)                    │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_16_project (Conv2D) │ (None, 7, 7, 320)      │        307,200 │ block_16_depthwise_re… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ block_16_project_BN       │ (None, 7, 7, 320)      │          1,280 │ block_16_project[0][0] │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ Conv_1 (Conv2D)           │ (None, 7, 7, 1280)     │        409,600 │ block_16_project_BN[0… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ Conv_1_bn                 │ (None, 7, 7, 1280)     │          5,120 │ Conv_1[0][0]           │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ out_relu (ReLU)           │ (None, 7, 7, 1280)     │              0 │ Conv_1_bn[0][0]        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ global_average_pooling2d… │ (None, 1280)           │              0 │ out_relu[0][0]         │
│ (GlobalAveragePooling2D)  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ dense_10 (Dense)          │ (None, 256)            │        327,936 │ global_average_poolin… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ dropout_5 (Dropout)       │ (None, 256)            │              0 │ dense_10[0][0]         │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ dense_11 (Dense)          │ (None, 1)              │            257 │ dropout_5[0][0]        │
└───────────────────────────┴────────────────────────┴────────────────┴────────────────────────┘

 Total params: 2,586,177 (9.87 MB)

 Trainable params: 328,193 (1.25 MB)

 Non-trainable params: 2,257,984 (8.61 MB)

# Early stopping
early_stopping = EarlyStopping(monitor='val_loss', patience=5, restore_best_weights=True)

# Start timer
start_time = time.time()

# Train the model
history = mobilenet_model.fit(
    train_generator,
    epochs=20,
    validation_data=test_generator,
    callbacks=[early_stopping]
)

# End timer
end_time = time.time()

# Calculate training duration
training_time = end_time - start_time
minutes, seconds = divmod(training_time, 60)
print("-" * 120)
print(f"Training completed in {int(minutes)} minutes and {int(seconds)} seconds.")

# Plot accuracy and loss graphs
plt.figure(figsize=(12, 3))

# Plot Accuracy
plt.subplot(1, 2, 1)
plt.plot(history.history['accuracy'], label='Train Accuracy')
plt.plot(history.history['val_accuracy'], label='Test Accuracy')
plt.xlabel('Epochs')
plt.ylabel('Accuracy')
plt.legend()
plt.title('\nTraining vs Test Accuracy\n')

# Plot Loss
plt.subplot(1, 2, 2)
plt.plot(history.history['loss'], label='Train Loss')
plt.plot(history.history['val_loss'], label='Test Loss')
plt.xlabel('Epochs')
plt.ylabel('Loss')
plt.legend()
plt.title('\nTraining vs Test Loss\n')

plt.show()

Epoch 1/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 23s 97ms/step - accuracy: 0.7379 - loss: 0.5367 - val_accuracy: 0.8606 - val_loss: 0.3143
Epoch 2/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 11s 62ms/step - accuracy: 0.8490 - loss: 0.3196 - val_accuracy: 0.8913 - val_loss: 0.2602
Epoch 3/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 10s 61ms/step - accuracy: 0.8870 - loss: 0.2423 - val_accuracy: 0.9234 - val_loss: 0.1957
Epoch 4/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 10s 60ms/step - accuracy: 0.9064 - loss: 0.2116 - val_accuracy: 0.9219 - val_loss: 0.1774
Epoch 5/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 10s 62ms/step - accuracy: 0.9165 - loss: 0.1871 - val_accuracy: 0.9357 - val_loss: 0.1549
Epoch 6/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 11s 63ms/step - accuracy: 0.9307 - loss: 0.1614 - val_accuracy: 0.9433 - val_loss: 0.1417
Epoch 7/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 10s 61ms/step - accuracy: 0.9406 - loss: 0.1377 - val_accuracy: 0.9495 - val_loss: 0.1241
Epoch 8/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 10s 61ms/step - accuracy: 0.9639 - loss: 0.1048 - val_accuracy: 0.9449 - val_loss: 0.1339
Epoch 9/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 10s 61ms/step - accuracy: 0.9584 - loss: 0.0988 - val_accuracy: 0.9556 - val_loss: 0.1162
Epoch 10/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 10s 58ms/step - accuracy: 0.9595 - loss: 0.1004 - val_accuracy: 0.9525 - val_loss: 0.1112
Epoch 11/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 10s 61ms/step - accuracy: 0.9598 - loss: 0.1135 - val_accuracy: 0.9587 - val_loss: 0.0993
Epoch 12/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 10s 61ms/step - accuracy: 0.9651 - loss: 0.0950 - val_accuracy: 0.9541 - val_loss: 0.1053
Epoch 13/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 10s 60ms/step - accuracy: 0.9774 - loss: 0.0699 - val_accuracy: 0.9587 - val_loss: 0.0882
Epoch 14/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 10s 60ms/step - accuracy: 0.9657 - loss: 0.0882 - val_accuracy: 0.9724 - val_loss: 0.0775
Epoch 15/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 10s 61ms/step - accuracy: 0.9793 - loss: 0.0539 - val_accuracy: 0.9571 - val_loss: 0.0977
Epoch 16/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 10s 60ms/step - accuracy: 0.9829 - loss: 0.0504 - val_accuracy: 0.9724 - val_loss: 0.0695
Epoch 17/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 10s 59ms/step - accuracy: 0.9851 - loss: 0.0478 - val_accuracy: 0.9694 - val_loss: 0.0909
Epoch 18/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 10s 61ms/step - accuracy: 0.9788 - loss: 0.0633 - val_accuracy: 0.9587 - val_loss: 0.0915
Epoch 19/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 10s 59ms/step - accuracy: 0.9926 - loss: 0.0300 - val_accuracy: 0.9724 - val_loss: 0.0747
Epoch 20/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 10s 59ms/step - accuracy: 0.9849 - loss: 0.0444 - val_accuracy: 0.9709 - val_loss: 0.0805
------------------------------------------------------------------------------------------------------------------------
Training completed in 3 minutes and 39 seconds.

# Predict the test set
y_pred_prob = mobilenet_model.predict(test_generator)
y_pred = (y_pred_prob > 0.5).astype(int)

# Get true labels
y_true = test_generator.classes

# Confusion Matrix
conf_matrix = confusion_matrix(y_true, y_pred)
plt.figure(figsize=(4, 3))
sns.heatmap(conf_matrix, annot=True, fmt='d', cmap='Blues', 
            xticklabels=['Healthy', 'Parkinson'], 
            yticklabels=['Healthy', 'Parkinson'], 
            cbar=False)
plt.xlabel('Predicted')
plt.ylabel('Actual')
plt.title('\nTest Data\n')
plt.show()

# Evaluation Metrics
accuracy = accuracy_score(y_true, y_pred)
f1 = f1_score(y_true, y_pred)
auc = roc_auc_score(y_true, y_pred_prob)

print(f'Accuracy: {accuracy:.2f}')
print(f'F1 Score: {f1:.2f}')
print(f'AUC Score: {auc:.2f}')

41/41 ━━━━━━━━━━━━━━━━━━━━ 2s 50ms/step

Accuracy: 0.97
F1 Score: 0.97
AUC Score: 1.00

---

3. VGG16:

• The architecture shares common elements with MobileNetV2, including GlobalAveragePooling2D, Dropout (0.5), and a Dense (Sigmoid) output layer for binary classification.

• Key Difference: The fully connected layer uses 512 neurons (vs. 256 in MobileNetV2) to accommodate VGG16's high-dimensional feature maps (e.g., 7x7x512), ensuring it captures the complex features effectively.

from tensorflow.keras.applications import VGG16

# Load VGG16 as the base model (excluding the top classification layers)
base_model = VGG16(weights='imagenet', include_top=False, input_shape=(224, 224, 3))

# Freeze all layers of the VGG16 pre-trained model
base_model.trainable = False

# Create custom classification layers
x = base_model.output
x = GlobalAveragePooling2D()(x)
x = Dense(512, activation='relu')(x)
x = Dropout(0.5)(x)
output = Dense(1, activation='sigmoid')(x)

# Build the complete model
vgg_model = Model(inputs=base_model.input, outputs=output)

# Compile the model
vgg_model.compile(optimizer='adam', loss='binary_crossentropy', metrics=['accuracy'])

# Model summary
vgg_model.summary()

Model: "functional_7"

┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━┓
┃ Layer (type)                         ┃ Output Shape                ┃         Param # ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━┩
│ input_layer_8 (InputLayer)           │ (None, 224, 224, 3)         │               0 │
├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤
│ block1_conv1 (Conv2D)                │ (None, 224, 224, 64)        │           1,792 │
├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤
│ block1_conv2 (Conv2D)                │ (None, 224, 224, 64)        │          36,928 │
├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤
│ block1_pool (MaxPooling2D)           │ (None, 112, 112, 64)        │               0 │
├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤
│ block2_conv1 (Conv2D)                │ (None, 112, 112, 128)       │          73,856 │
├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤
│ block2_conv2 (Conv2D)                │ (None, 112, 112, 128)       │         147,584 │
├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤
│ block2_pool (MaxPooling2D)           │ (None, 56, 56, 128)         │               0 │
├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤
│ block3_conv1 (Conv2D)                │ (None, 56, 56, 256)         │         295,168 │
├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤
│ block3_conv2 (Conv2D)                │ (None, 56, 56, 256)         │         590,080 │
├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤
│ block3_conv3 (Conv2D)                │ (None, 56, 56, 256)         │         590,080 │
├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤
│ block3_pool (MaxPooling2D)           │ (None, 28, 28, 256)         │               0 │
├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤
│ block4_conv1 (Conv2D)                │ (None, 28, 28, 512)         │       1,180,160 │
├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤
│ block4_conv2 (Conv2D)                │ (None, 28, 28, 512)         │       2,359,808 │
├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤
│ block4_conv3 (Conv2D)                │ (None, 28, 28, 512)         │       2,359,808 │
├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤
│ block4_pool (MaxPooling2D)           │ (None, 14, 14, 512)         │               0 │
├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤
│ block5_conv1 (Conv2D)                │ (None, 14, 14, 512)         │       2,359,808 │
├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤
│ block5_conv2 (Conv2D)                │ (None, 14, 14, 512)         │       2,359,808 │
├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤
│ block5_conv3 (Conv2D)                │ (None, 14, 14, 512)         │       2,359,808 │
├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤
│ block5_pool (MaxPooling2D)           │ (None, 7, 7, 512)           │               0 │
├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤
│ global_average_pooling2d_4           │ (None, 512)                 │               0 │
│ (GlobalAveragePooling2D)             │                             │                 │
├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤
│ dense_14 (Dense)                     │ (None, 512)                 │         262,656 │
├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤
│ dropout_7 (Dropout)                  │ (None, 512)                 │               0 │
├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤
│ dense_15 (Dense)                     │ (None, 1)                   │             513 │
└──────────────────────────────────────┴─────────────────────────────┴─────────────────┘

 Total params: 14,977,857 (57.14 MB)

 Trainable params: 263,169 (1.00 MB)

 Non-trainable params: 14,714,688 (56.13 MB)

# Early stopping
early_stopping = EarlyStopping(monitor='val_loss', patience=5, restore_best_weights=True)

# Start timer
start_time = time.time()

# Train the model
history = vgg_model.fit(
    train_generator,
    epochs=20,
    validation_data=test_generator,
    callbacks=[early_stopping]
)

# End timer
end_time = time.time()

# Calculate training duration
training_time = end_time - start_time
minutes, seconds = divmod(training_time, 60)
print("-" * 120)
print(f"Training completed in {int(minutes)} minutes and {int(seconds)} seconds.")

# Plot accuracy and loss graphs
plt.figure(figsize=(12, 3))

# Plot Accuracy
plt.subplot(1, 2, 1)
plt.plot(history.history['accuracy'], label='Train Accuracy')
plt.plot(history.history['val_accuracy'], label='Test Accuracy')
plt.xlabel('Epochs')
plt.ylabel('Accuracy')
plt.legend()
plt.title('\nTraining vs Test Accuracy\n')

# Plot Loss
plt.subplot(1, 2, 2)
plt.plot(history.history['loss'], label='Train Loss')
plt.plot(history.history['val_loss'], label='Test Loss')
plt.xlabel('Epochs')
plt.ylabel('Loss')
plt.legend()
plt.title('\nTraining vs Test Loss\n')

plt.show()

Epoch 1/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 19s 98ms/step - accuracy: 0.6762 - loss: 0.5997 - val_accuracy: 0.8484 - val_loss: 0.3751
Epoch 2/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 15s 88ms/step - accuracy: 0.8029 - loss: 0.4254 - val_accuracy: 0.8315 - val_loss: 0.3775
Epoch 3/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 14s 85ms/step - accuracy: 0.8215 - loss: 0.3965 - val_accuracy: 0.8560 - val_loss: 0.3371
Epoch 4/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 14s 83ms/step - accuracy: 0.8357 - loss: 0.3695 - val_accuracy: 0.8729 - val_loss: 0.3030
Epoch 5/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 14s 82ms/step - accuracy: 0.8602 - loss: 0.3338 - val_accuracy: 0.8744 - val_loss: 0.2837
Epoch 6/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 14s 83ms/step - accuracy: 0.8609 - loss: 0.3176 - val_accuracy: 0.8821 - val_loss: 0.2862
Epoch 7/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 14s 84ms/step - accuracy: 0.8528 - loss: 0.3287 - val_accuracy: 0.8836 - val_loss: 0.2617
Epoch 8/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 14s 84ms/step - accuracy: 0.8630 - loss: 0.2953 - val_accuracy: 0.8760 - val_loss: 0.2625
Epoch 9/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 14s 84ms/step - accuracy: 0.8689 - loss: 0.2969 - val_accuracy: 0.8897 - val_loss: 0.2639
Epoch 10/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 14s 84ms/step - accuracy: 0.8608 - loss: 0.2952 - val_accuracy: 0.8882 - val_loss: 0.2493
Epoch 11/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 14s 83ms/step - accuracy: 0.8801 - loss: 0.2619 - val_accuracy: 0.8821 - val_loss: 0.2519
Epoch 12/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 14s 83ms/step - accuracy: 0.8816 - loss: 0.2725 - val_accuracy: 0.9035 - val_loss: 0.2201
Epoch 13/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 14s 83ms/step - accuracy: 0.8755 - loss: 0.2927 - val_accuracy: 0.9035 - val_loss: 0.2105
Epoch 14/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 14s 84ms/step - accuracy: 0.8989 - loss: 0.2320 - val_accuracy: 0.9066 - val_loss: 0.2091
Epoch 15/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 14s 83ms/step - accuracy: 0.9115 - loss: 0.2089 - val_accuracy: 0.9142 - val_loss: 0.2127
Epoch 16/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 14s 84ms/step - accuracy: 0.8984 - loss: 0.2246 - val_accuracy: 0.9096 - val_loss: 0.2018
Epoch 17/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 14s 84ms/step - accuracy: 0.8998 - loss: 0.2349 - val_accuracy: 0.9204 - val_loss: 0.1927
Epoch 18/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 14s 84ms/step - accuracy: 0.9136 - loss: 0.2059 - val_accuracy: 0.9158 - val_loss: 0.1886
Epoch 19/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 14s 84ms/step - accuracy: 0.9150 - loss: 0.1989 - val_accuracy: 0.9188 - val_loss: 0.1911
Epoch 20/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 14s 84ms/step - accuracy: 0.9191 - loss: 0.1971 - val_accuracy: 0.9173 - val_loss: 0.1949
------------------------------------------------------------------------------------------------------------------------
Training completed in 4 minutes and 47 seconds.

# Predict the test set
y_pred_prob = vgg_model.predict(test_generator)
y_pred = (y_pred_prob > 0.5).astype(int)

# Get true labels
y_true = test_generator.classes

# Confusion Matrix
conf_matrix = confusion_matrix(y_true, y_pred)
plt.figure(figsize=(4, 3))
sns.heatmap(conf_matrix, annot=True, fmt='d', cmap='Blues', 
            xticklabels=['Healthy', 'Parkinson'], 
            yticklabels=['Healthy', 'Parkinson'], 
            cbar=False)
plt.xlabel('Predicted')
plt.ylabel('Actual')
plt.title('\nTest Data\n')
plt.show()

# Evaluation Metrics
accuracy = accuracy_score(y_true, y_pred)
f1 = f1_score(y_true, y_pred)
auc = roc_auc_score(y_true, y_pred_prob)

print(f'Accuracy: {accuracy:.2f}')
print(f'F1 Score: {f1:.2f}')
print(f'AUC Score: {auc:.2f}')

41/41 ━━━━━━━━━━━━━━━━━━━━ 3s 73ms/step

Accuracy: 0.92
F1 Score: 0.91
AUC Score: 0.98

---

4. DenseNet121:

• The architecture shares common elements with MobileNetV2 and VGG16, including GlobalAveragePooling2D, Dropout (0.5), and a Dense (Sigmoid) output layer for binary classification.

• Key Difference: DenseNet121’s densely connected layers promote feature reuse and efficient gradient flow, producing richer and more compact feature maps compared to VGG16’s independent layer design. The 512-neuron Dense layer ensures these complex outputs are effectively utilized without bottlenecking.

from tensorflow.keras.applications import DenseNet121

# Load DenseNet121 as the base model (excluding the top classification layers)
base_model = DenseNet121(weights='imagenet', include_top=False, input_shape=(224, 224, 3))

# Freeze all layers of the DenseNet121 pre-trained model
base_model.trainable = False

# Create custom classification layers
x = base_model.output
x = GlobalAveragePooling2D()(x)
x = Dense(512, activation='relu')(x)
x = Dropout(0.5)(x)
output = Dense(1, activation='sigmoid')(x)

# Build the complete model
densenet_model = Model(inputs=base_model.input, outputs=output)

# Compile the model
densenet_model.compile(optimizer='adam', loss='binary_crossentropy', metrics=['accuracy'])

# Model summary
densenet_model.summary()

Model: "functional_8"

┏━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━┓
┃ Layer (type)              ┃ Output Shape           ┃        Param # ┃ Connected to           ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━┩
│ input_layer_9             │ (None, 224, 224, 3)    │              0 │ -                      │
│ (InputLayer)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ zero_padding2d_2          │ (None, 230, 230, 3)    │              0 │ input_layer_9[0][0]    │
│ (ZeroPadding2D)           │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv1_conv (Conv2D)       │ (None, 112, 112, 64)   │          9,408 │ zero_padding2d_2[0][0] │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv1_bn                  │ (None, 112, 112, 64)   │            256 │ conv1_conv[0][0]       │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv1_relu (Activation)   │ (None, 112, 112, 64)   │              0 │ conv1_bn[0][0]         │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ zero_padding2d_3          │ (None, 114, 114, 64)   │              0 │ conv1_relu[0][0]       │
│ (ZeroPadding2D)           │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ pool1 (MaxPooling2D)      │ (None, 56, 56, 64)     │              0 │ zero_padding2d_3[0][0] │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv2_block1_0_bn         │ (None, 56, 56, 64)     │            256 │ pool1[0][0]            │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv2_block1_0_relu       │ (None, 56, 56, 64)     │              0 │ conv2_block1_0_bn[0][… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv2_block1_1_conv       │ (None, 56, 56, 128)    │          8,192 │ conv2_block1_0_relu[0… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv2_block1_1_bn         │ (None, 56, 56, 128)    │            512 │ conv2_block1_1_conv[0… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv2_block1_1_relu       │ (None, 56, 56, 128)    │              0 │ conv2_block1_1_bn[0][… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv2_block1_2_conv       │ (None, 56, 56, 32)     │         36,864 │ conv2_block1_1_relu[0… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv2_block1_concat       │ (None, 56, 56, 96)     │              0 │ pool1[0][0],           │
│ (Concatenate)             │                        │                │ conv2_block1_2_conv[0… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv2_block2_0_bn         │ (None, 56, 56, 96)     │            384 │ conv2_block1_concat[0… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv2_block2_0_relu       │ (None, 56, 56, 96)     │              0 │ conv2_block2_0_bn[0][… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv2_block2_1_conv       │ (None, 56, 56, 128)    │         12,288 │ conv2_block2_0_relu[0… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv2_block2_1_bn         │ (None, 56, 56, 128)    │            512 │ conv2_block2_1_conv[0… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv2_block2_1_relu       │ (None, 56, 56, 128)    │              0 │ conv2_block2_1_bn[0][… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv2_block2_2_conv       │ (None, 56, 56, 32)     │         36,864 │ conv2_block2_1_relu[0… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv2_block2_concat       │ (None, 56, 56, 128)    │              0 │ conv2_block1_concat[0… │
│ (Concatenate)             │                        │                │ conv2_block2_2_conv[0… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv2_block3_0_bn         │ (None, 56, 56, 128)    │            512 │ conv2_block2_concat[0… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv2_block3_0_relu       │ (None, 56, 56, 128)    │              0 │ conv2_block3_0_bn[0][… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv2_block3_1_conv       │ (None, 56, 56, 128)    │         16,384 │ conv2_block3_0_relu[0… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv2_block3_1_bn         │ (None, 56, 56, 128)    │            512 │ conv2_block3_1_conv[0… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv2_block3_1_relu       │ (None, 56, 56, 128)    │              0 │ conv2_block3_1_bn[0][… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv2_block3_2_conv       │ (None, 56, 56, 32)     │         36,864 │ conv2_block3_1_relu[0… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv2_block3_concat       │ (None, 56, 56, 160)    │              0 │ conv2_block2_concat[0… │
│ (Concatenate)             │                        │                │ conv2_block3_2_conv[0… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv2_block4_0_bn         │ (None, 56, 56, 160)    │            640 │ conv2_block3_concat[0… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv2_block4_0_relu       │ (None, 56, 56, 160)    │              0 │ conv2_block4_0_bn[0][… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv2_block4_1_conv       │ (None, 56, 56, 128)    │         20,480 │ conv2_block4_0_relu[0… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv2_block4_1_bn         │ (None, 56, 56, 128)    │            512 │ conv2_block4_1_conv[0… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv2_block4_1_relu       │ (None, 56, 56, 128)    │              0 │ conv2_block4_1_bn[0][… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv2_block4_2_conv       │ (None, 56, 56, 32)     │         36,864 │ conv2_block4_1_relu[0… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv2_block4_concat       │ (None, 56, 56, 192)    │              0 │ conv2_block3_concat[0… │
│ (Concatenate)             │                        │                │ conv2_block4_2_conv[0… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv2_block5_0_bn         │ (None, 56, 56, 192)    │            768 │ conv2_block4_concat[0… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv2_block5_0_relu       │ (None, 56, 56, 192)    │              0 │ conv2_block5_0_bn[0][… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv2_block5_1_conv       │ (None, 56, 56, 128)    │         24,576 │ conv2_block5_0_relu[0… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv2_block5_1_bn         │ (None, 56, 56, 128)    │            512 │ conv2_block5_1_conv[0… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv2_block5_1_relu       │ (None, 56, 56, 128)    │              0 │ conv2_block5_1_bn[0][… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv2_block5_2_conv       │ (None, 56, 56, 32)     │         36,864 │ conv2_block5_1_relu[0… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv2_block5_concat       │ (None, 56, 56, 224)    │              0 │ conv2_block4_concat[0… │
│ (Concatenate)             │                        │                │ conv2_block5_2_conv[0… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv2_block6_0_bn         │ (None, 56, 56, 224)    │            896 │ conv2_block5_concat[0… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv2_block6_0_relu       │ (None, 56, 56, 224)    │              0 │ conv2_block6_0_bn[0][… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv2_block6_1_conv       │ (None, 56, 56, 128)    │         28,672 │ conv2_block6_0_relu[0… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv2_block6_1_bn         │ (None, 56, 56, 128)    │            512 │ conv2_block6_1_conv[0… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv2_block6_1_relu       │ (None, 56, 56, 128)    │              0 │ conv2_block6_1_bn[0][… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv2_block6_2_conv       │ (None, 56, 56, 32)     │         36,864 │ conv2_block6_1_relu[0… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv2_block6_concat       │ (None, 56, 56, 256)    │              0 │ conv2_block5_concat[0… │
│ (Concatenate)             │                        │                │ conv2_block6_2_conv[0… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ pool2_bn                  │ (None, 56, 56, 256)    │          1,024 │ conv2_block6_concat[0… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ pool2_relu (Activation)   │ (None, 56, 56, 256)    │              0 │ pool2_bn[0][0]         │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ pool2_conv (Conv2D)       │ (None, 56, 56, 128)    │         32,768 │ pool2_relu[0][0]       │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ pool2_pool                │ (None, 28, 28, 128)    │              0 │ pool2_conv[0][0]       │
│ (AveragePooling2D)        │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block1_0_bn         │ (None, 28, 28, 128)    │            512 │ pool2_pool[0][0]       │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block1_0_relu       │ (None, 28, 28, 128)    │              0 │ conv3_block1_0_bn[0][… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block1_1_conv       │ (None, 28, 28, 128)    │         16,384 │ conv3_block1_0_relu[0… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block1_1_bn         │ (None, 28, 28, 128)    │            512 │ conv3_block1_1_conv[0… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block1_1_relu       │ (None, 28, 28, 128)    │              0 │ conv3_block1_1_bn[0][… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block1_2_conv       │ (None, 28, 28, 32)     │         36,864 │ conv3_block1_1_relu[0… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block1_concat       │ (None, 28, 28, 160)    │              0 │ pool2_pool[0][0],      │
│ (Concatenate)             │                        │                │ conv3_block1_2_conv[0… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block2_0_bn         │ (None, 28, 28, 160)    │            640 │ conv3_block1_concat[0… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block2_0_relu       │ (None, 28, 28, 160)    │              0 │ conv3_block2_0_bn[0][… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block2_1_conv       │ (None, 28, 28, 128)    │         20,480 │ conv3_block2_0_relu[0… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block2_1_bn         │ (None, 28, 28, 128)    │            512 │ conv3_block2_1_conv[0… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block2_1_relu       │ (None, 28, 28, 128)    │              0 │ conv3_block2_1_bn[0][… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block2_2_conv       │ (None, 28, 28, 32)     │         36,864 │ conv3_block2_1_relu[0… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block2_concat       │ (None, 28, 28, 192)    │              0 │ conv3_block1_concat[0… │
│ (Concatenate)             │                        │                │ conv3_block2_2_conv[0… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block3_0_bn         │ (None, 28, 28, 192)    │            768 │ conv3_block2_concat[0… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block3_0_relu       │ (None, 28, 28, 192)    │              0 │ conv3_block3_0_bn[0][… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block3_1_conv       │ (None, 28, 28, 128)    │         24,576 │ conv3_block3_0_relu[0… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block3_1_bn         │ (None, 28, 28, 128)    │            512 │ conv3_block3_1_conv[0… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block3_1_relu       │ (None, 28, 28, 128)    │              0 │ conv3_block3_1_bn[0][… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block3_2_conv       │ (None, 28, 28, 32)     │         36,864 │ conv3_block3_1_relu[0… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block3_concat       │ (None, 28, 28, 224)    │              0 │ conv3_block2_concat[0… │
│ (Concatenate)             │                        │                │ conv3_block3_2_conv[0… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block4_0_bn         │ (None, 28, 28, 224)    │            896 │ conv3_block3_concat[0… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block4_0_relu       │ (None, 28, 28, 224)    │              0 │ conv3_block4_0_bn[0][… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block4_1_conv       │ (None, 28, 28, 128)    │         28,672 │ conv3_block4_0_relu[0… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block4_1_bn         │ (None, 28, 28, 128)    │            512 │ conv3_block4_1_conv[0… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block4_1_relu       │ (None, 28, 28, 128)    │              0 │ conv3_block4_1_bn[0][… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block4_2_conv       │ (None, 28, 28, 32)     │         36,864 │ conv3_block4_1_relu[0… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block4_concat       │ (None, 28, 28, 256)    │              0 │ conv3_block3_concat[0… │
│ (Concatenate)             │                        │                │ conv3_block4_2_conv[0… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block5_0_bn         │ (None, 28, 28, 256)    │          1,024 │ conv3_block4_concat[0… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block5_0_relu       │ (None, 28, 28, 256)    │              0 │ conv3_block5_0_bn[0][… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block5_1_conv       │ (None, 28, 28, 128)    │         32,768 │ conv3_block5_0_relu[0… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block5_1_bn         │ (None, 28, 28, 128)    │            512 │ conv3_block5_1_conv[0… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block5_1_relu       │ (None, 28, 28, 128)    │              0 │ conv3_block5_1_bn[0][… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block5_2_conv       │ (None, 28, 28, 32)     │         36,864 │ conv3_block5_1_relu[0… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block5_concat       │ (None, 28, 28, 288)    │              0 │ conv3_block4_concat[0… │
│ (Concatenate)             │                        │                │ conv3_block5_2_conv[0… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block6_0_bn         │ (None, 28, 28, 288)    │          1,152 │ conv3_block5_concat[0… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block6_0_relu       │ (None, 28, 28, 288)    │              0 │ conv3_block6_0_bn[0][… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block6_1_conv       │ (None, 28, 28, 128)    │         36,864 │ conv3_block6_0_relu[0… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block6_1_bn         │ (None, 28, 28, 128)    │            512 │ conv3_block6_1_conv[0… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block6_1_relu       │ (None, 28, 28, 128)    │              0 │ conv3_block6_1_bn[0][… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block6_2_conv       │ (None, 28, 28, 32)     │         36,864 │ conv3_block6_1_relu[0… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block6_concat       │ (None, 28, 28, 320)    │              0 │ conv3_block5_concat[0… │
│ (Concatenate)             │                        │                │ conv3_block6_2_conv[0… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block7_0_bn         │ (None, 28, 28, 320)    │          1,280 │ conv3_block6_concat[0… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block7_0_relu       │ (None, 28, 28, 320)    │              0 │ conv3_block7_0_bn[0][… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block7_1_conv       │ (None, 28, 28, 128)    │         40,960 │ conv3_block7_0_relu[0… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block7_1_bn         │ (None, 28, 28, 128)    │            512 │ conv3_block7_1_conv[0… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block7_1_relu       │ (None, 28, 28, 128)    │              0 │ conv3_block7_1_bn[0][… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block7_2_conv       │ (None, 28, 28, 32)     │         36,864 │ conv3_block7_1_relu[0… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block7_concat       │ (None, 28, 28, 352)    │              0 │ conv3_block6_concat[0… │
│ (Concatenate)             │                        │                │ conv3_block7_2_conv[0… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block8_0_bn         │ (None, 28, 28, 352)    │          1,408 │ conv3_block7_concat[0… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block8_0_relu       │ (None, 28, 28, 352)    │              0 │ conv3_block8_0_bn[0][… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block8_1_conv       │ (None, 28, 28, 128)    │         45,056 │ conv3_block8_0_relu[0… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block8_1_bn         │ (None, 28, 28, 128)    │            512 │ conv3_block8_1_conv[0… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block8_1_relu       │ (None, 28, 28, 128)    │              0 │ conv3_block8_1_bn[0][… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block8_2_conv       │ (None, 28, 28, 32)     │         36,864 │ conv3_block8_1_relu[0… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block8_concat       │ (None, 28, 28, 384)    │              0 │ conv3_block7_concat[0… │
│ (Concatenate)             │                        │                │ conv3_block8_2_conv[0… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block9_0_bn         │ (None, 28, 28, 384)    │          1,536 │ conv3_block8_concat[0… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block9_0_relu       │ (None, 28, 28, 384)    │              0 │ conv3_block9_0_bn[0][… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block9_1_conv       │ (None, 28, 28, 128)    │         49,152 │ conv3_block9_0_relu[0… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block9_1_bn         │ (None, 28, 28, 128)    │            512 │ conv3_block9_1_conv[0… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block9_1_relu       │ (None, 28, 28, 128)    │              0 │ conv3_block9_1_bn[0][… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block9_2_conv       │ (None, 28, 28, 32)     │         36,864 │ conv3_block9_1_relu[0… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block9_concat       │ (None, 28, 28, 416)    │              0 │ conv3_block8_concat[0… │
│ (Concatenate)             │                        │                │ conv3_block9_2_conv[0… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block10_0_bn        │ (None, 28, 28, 416)    │          1,664 │ conv3_block9_concat[0… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block10_0_relu      │ (None, 28, 28, 416)    │              0 │ conv3_block10_0_bn[0]… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block10_1_conv      │ (None, 28, 28, 128)    │         53,248 │ conv3_block10_0_relu[… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block10_1_bn        │ (None, 28, 28, 128)    │            512 │ conv3_block10_1_conv[… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block10_1_relu      │ (None, 28, 28, 128)    │              0 │ conv3_block10_1_bn[0]… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block10_2_conv      │ (None, 28, 28, 32)     │         36,864 │ conv3_block10_1_relu[… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block10_concat      │ (None, 28, 28, 448)    │              0 │ conv3_block9_concat[0… │
│ (Concatenate)             │                        │                │ conv3_block10_2_conv[… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block11_0_bn        │ (None, 28, 28, 448)    │          1,792 │ conv3_block10_concat[… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block11_0_relu      │ (None, 28, 28, 448)    │              0 │ conv3_block11_0_bn[0]… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block11_1_conv      │ (None, 28, 28, 128)    │         57,344 │ conv3_block11_0_relu[… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block11_1_bn        │ (None, 28, 28, 128)    │            512 │ conv3_block11_1_conv[… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block11_1_relu      │ (None, 28, 28, 128)    │              0 │ conv3_block11_1_bn[0]… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block11_2_conv      │ (None, 28, 28, 32)     │         36,864 │ conv3_block11_1_relu[… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block11_concat      │ (None, 28, 28, 480)    │              0 │ conv3_block10_concat[… │
│ (Concatenate)             │                        │                │ conv3_block11_2_conv[… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block12_0_bn        │ (None, 28, 28, 480)    │          1,920 │ conv3_block11_concat[… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block12_0_relu      │ (None, 28, 28, 480)    │              0 │ conv3_block12_0_bn[0]… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block12_1_conv      │ (None, 28, 28, 128)    │         61,440 │ conv3_block12_0_relu[… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block12_1_bn        │ (None, 28, 28, 128)    │            512 │ conv3_block12_1_conv[… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block12_1_relu      │ (None, 28, 28, 128)    │              0 │ conv3_block12_1_bn[0]… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block12_2_conv      │ (None, 28, 28, 32)     │         36,864 │ conv3_block12_1_relu[… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv3_block12_concat      │ (None, 28, 28, 512)    │              0 │ conv3_block11_concat[… │
│ (Concatenate)             │                        │                │ conv3_block12_2_conv[… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ pool3_bn                  │ (None, 28, 28, 512)    │          2,048 │ conv3_block12_concat[… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ pool3_relu (Activation)   │ (None, 28, 28, 512)    │              0 │ pool3_bn[0][0]         │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ pool3_conv (Conv2D)       │ (None, 28, 28, 256)    │        131,072 │ pool3_relu[0][0]       │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ pool3_pool                │ (None, 14, 14, 256)    │              0 │ pool3_conv[0][0]       │
│ (AveragePooling2D)        │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block1_0_bn         │ (None, 14, 14, 256)    │          1,024 │ pool3_pool[0][0]       │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block1_0_relu       │ (None, 14, 14, 256)    │              0 │ conv4_block1_0_bn[0][… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block1_1_conv       │ (None, 14, 14, 128)    │         32,768 │ conv4_block1_0_relu[0… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block1_1_bn         │ (None, 14, 14, 128)    │            512 │ conv4_block1_1_conv[0… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block1_1_relu       │ (None, 14, 14, 128)    │              0 │ conv4_block1_1_bn[0][… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block1_2_conv       │ (None, 14, 14, 32)     │         36,864 │ conv4_block1_1_relu[0… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block1_concat       │ (None, 14, 14, 288)    │              0 │ pool3_pool[0][0],      │
│ (Concatenate)             │                        │                │ conv4_block1_2_conv[0… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block2_0_bn         │ (None, 14, 14, 288)    │          1,152 │ conv4_block1_concat[0… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block2_0_relu       │ (None, 14, 14, 288)    │              0 │ conv4_block2_0_bn[0][… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block2_1_conv       │ (None, 14, 14, 128)    │         36,864 │ conv4_block2_0_relu[0… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block2_1_bn         │ (None, 14, 14, 128)    │            512 │ conv4_block2_1_conv[0… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block2_1_relu       │ (None, 14, 14, 128)    │              0 │ conv4_block2_1_bn[0][… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block2_2_conv       │ (None, 14, 14, 32)     │         36,864 │ conv4_block2_1_relu[0… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block2_concat       │ (None, 14, 14, 320)    │              0 │ conv4_block1_concat[0… │
│ (Concatenate)             │                        │                │ conv4_block2_2_conv[0… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block3_0_bn         │ (None, 14, 14, 320)    │          1,280 │ conv4_block2_concat[0… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block3_0_relu       │ (None, 14, 14, 320)    │              0 │ conv4_block3_0_bn[0][… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block3_1_conv       │ (None, 14, 14, 128)    │         40,960 │ conv4_block3_0_relu[0… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block3_1_bn         │ (None, 14, 14, 128)    │            512 │ conv4_block3_1_conv[0… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block3_1_relu       │ (None, 14, 14, 128)    │              0 │ conv4_block3_1_bn[0][… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block3_2_conv       │ (None, 14, 14, 32)     │         36,864 │ conv4_block3_1_relu[0… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block3_concat       │ (None, 14, 14, 352)    │              0 │ conv4_block2_concat[0… │
│ (Concatenate)             │                        │                │ conv4_block3_2_conv[0… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block4_0_bn         │ (None, 14, 14, 352)    │          1,408 │ conv4_block3_concat[0… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block4_0_relu       │ (None, 14, 14, 352)    │              0 │ conv4_block4_0_bn[0][… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block4_1_conv       │ (None, 14, 14, 128)    │         45,056 │ conv4_block4_0_relu[0… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block4_1_bn         │ (None, 14, 14, 128)    │            512 │ conv4_block4_1_conv[0… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block4_1_relu       │ (None, 14, 14, 128)    │              0 │ conv4_block4_1_bn[0][… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block4_2_conv       │ (None, 14, 14, 32)     │         36,864 │ conv4_block4_1_relu[0… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block4_concat       │ (None, 14, 14, 384)    │              0 │ conv4_block3_concat[0… │
│ (Concatenate)             │                        │                │ conv4_block4_2_conv[0… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block5_0_bn         │ (None, 14, 14, 384)    │          1,536 │ conv4_block4_concat[0… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block5_0_relu       │ (None, 14, 14, 384)    │              0 │ conv4_block5_0_bn[0][… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block5_1_conv       │ (None, 14, 14, 128)    │         49,152 │ conv4_block5_0_relu[0… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block5_1_bn         │ (None, 14, 14, 128)    │            512 │ conv4_block5_1_conv[0… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block5_1_relu       │ (None, 14, 14, 128)    │              0 │ conv4_block5_1_bn[0][… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block5_2_conv       │ (None, 14, 14, 32)     │         36,864 │ conv4_block5_1_relu[0… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block5_concat       │ (None, 14, 14, 416)    │              0 │ conv4_block4_concat[0… │
│ (Concatenate)             │                        │                │ conv4_block5_2_conv[0… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block6_0_bn         │ (None, 14, 14, 416)    │          1,664 │ conv4_block5_concat[0… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block6_0_relu       │ (None, 14, 14, 416)    │              0 │ conv4_block6_0_bn[0][… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block6_1_conv       │ (None, 14, 14, 128)    │         53,248 │ conv4_block6_0_relu[0… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block6_1_bn         │ (None, 14, 14, 128)    │            512 │ conv4_block6_1_conv[0… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block6_1_relu       │ (None, 14, 14, 128)    │              0 │ conv4_block6_1_bn[0][… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block6_2_conv       │ (None, 14, 14, 32)     │         36,864 │ conv4_block6_1_relu[0… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block6_concat       │ (None, 14, 14, 448)    │              0 │ conv4_block5_concat[0… │
│ (Concatenate)             │                        │                │ conv4_block6_2_conv[0… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block7_0_bn         │ (None, 14, 14, 448)    │          1,792 │ conv4_block6_concat[0… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block7_0_relu       │ (None, 14, 14, 448)    │              0 │ conv4_block7_0_bn[0][… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block7_1_conv       │ (None, 14, 14, 128)    │         57,344 │ conv4_block7_0_relu[0… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block7_1_bn         │ (None, 14, 14, 128)    │            512 │ conv4_block7_1_conv[0… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block7_1_relu       │ (None, 14, 14, 128)    │              0 │ conv4_block7_1_bn[0][… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block7_2_conv       │ (None, 14, 14, 32)     │         36,864 │ conv4_block7_1_relu[0… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block7_concat       │ (None, 14, 14, 480)    │              0 │ conv4_block6_concat[0… │
│ (Concatenate)             │                        │                │ conv4_block7_2_conv[0… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block8_0_bn         │ (None, 14, 14, 480)    │          1,920 │ conv4_block7_concat[0… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block8_0_relu       │ (None, 14, 14, 480)    │              0 │ conv4_block8_0_bn[0][… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block8_1_conv       │ (None, 14, 14, 128)    │         61,440 │ conv4_block8_0_relu[0… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block8_1_bn         │ (None, 14, 14, 128)    │            512 │ conv4_block8_1_conv[0… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block8_1_relu       │ (None, 14, 14, 128)    │              0 │ conv4_block8_1_bn[0][… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block8_2_conv       │ (None, 14, 14, 32)     │         36,864 │ conv4_block8_1_relu[0… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block8_concat       │ (None, 14, 14, 512)    │              0 │ conv4_block7_concat[0… │
│ (Concatenate)             │                        │                │ conv4_block8_2_conv[0… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block9_0_bn         │ (None, 14, 14, 512)    │          2,048 │ conv4_block8_concat[0… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block9_0_relu       │ (None, 14, 14, 512)    │              0 │ conv4_block9_0_bn[0][… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block9_1_conv       │ (None, 14, 14, 128)    │         65,536 │ conv4_block9_0_relu[0… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block9_1_bn         │ (None, 14, 14, 128)    │            512 │ conv4_block9_1_conv[0… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block9_1_relu       │ (None, 14, 14, 128)    │              0 │ conv4_block9_1_bn[0][… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block9_2_conv       │ (None, 14, 14, 32)     │         36,864 │ conv4_block9_1_relu[0… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block9_concat       │ (None, 14, 14, 544)    │              0 │ conv4_block8_concat[0… │
│ (Concatenate)             │                        │                │ conv4_block9_2_conv[0… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block10_0_bn        │ (None, 14, 14, 544)    │          2,176 │ conv4_block9_concat[0… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block10_0_relu      │ (None, 14, 14, 544)    │              0 │ conv4_block10_0_bn[0]… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block10_1_conv      │ (None, 14, 14, 128)    │         69,632 │ conv4_block10_0_relu[… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block10_1_bn        │ (None, 14, 14, 128)    │            512 │ conv4_block10_1_conv[… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block10_1_relu      │ (None, 14, 14, 128)    │              0 │ conv4_block10_1_bn[0]… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block10_2_conv      │ (None, 14, 14, 32)     │         36,864 │ conv4_block10_1_relu[… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block10_concat      │ (None, 14, 14, 576)    │              0 │ conv4_block9_concat[0… │
│ (Concatenate)             │                        │                │ conv4_block10_2_conv[… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block11_0_bn        │ (None, 14, 14, 576)    │          2,304 │ conv4_block10_concat[… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block11_0_relu      │ (None, 14, 14, 576)    │              0 │ conv4_block11_0_bn[0]… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block11_1_conv      │ (None, 14, 14, 128)    │         73,728 │ conv4_block11_0_relu[… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block11_1_bn        │ (None, 14, 14, 128)    │            512 │ conv4_block11_1_conv[… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block11_1_relu      │ (None, 14, 14, 128)    │              0 │ conv4_block11_1_bn[0]… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block11_2_conv      │ (None, 14, 14, 32)     │         36,864 │ conv4_block11_1_relu[… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block11_concat      │ (None, 14, 14, 608)    │              0 │ conv4_block10_concat[… │
│ (Concatenate)             │                        │                │ conv4_block11_2_conv[… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block12_0_bn        │ (None, 14, 14, 608)    │          2,432 │ conv4_block11_concat[… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block12_0_relu      │ (None, 14, 14, 608)    │              0 │ conv4_block12_0_bn[0]… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block12_1_conv      │ (None, 14, 14, 128)    │         77,824 │ conv4_block12_0_relu[… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block12_1_bn        │ (None, 14, 14, 128)    │            512 │ conv4_block12_1_conv[… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block12_1_relu      │ (None, 14, 14, 128)    │              0 │ conv4_block12_1_bn[0]… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block12_2_conv      │ (None, 14, 14, 32)     │         36,864 │ conv4_block12_1_relu[… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block12_concat      │ (None, 14, 14, 640)    │              0 │ conv4_block11_concat[… │
│ (Concatenate)             │                        │                │ conv4_block12_2_conv[… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block13_0_bn        │ (None, 14, 14, 640)    │          2,560 │ conv4_block12_concat[… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block13_0_relu      │ (None, 14, 14, 640)    │              0 │ conv4_block13_0_bn[0]… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block13_1_conv      │ (None, 14, 14, 128)    │         81,920 │ conv4_block13_0_relu[… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block13_1_bn        │ (None, 14, 14, 128)    │            512 │ conv4_block13_1_conv[… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block13_1_relu      │ (None, 14, 14, 128)    │              0 │ conv4_block13_1_bn[0]… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block13_2_conv      │ (None, 14, 14, 32)     │         36,864 │ conv4_block13_1_relu[… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block13_concat      │ (None, 14, 14, 672)    │              0 │ conv4_block12_concat[… │
│ (Concatenate)             │                        │                │ conv4_block13_2_conv[… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block14_0_bn        │ (None, 14, 14, 672)    │          2,688 │ conv4_block13_concat[… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block14_0_relu      │ (None, 14, 14, 672)    │              0 │ conv4_block14_0_bn[0]… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block14_1_conv      │ (None, 14, 14, 128)    │         86,016 │ conv4_block14_0_relu[… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block14_1_bn        │ (None, 14, 14, 128)    │            512 │ conv4_block14_1_conv[… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block14_1_relu      │ (None, 14, 14, 128)    │              0 │ conv4_block14_1_bn[0]… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block14_2_conv      │ (None, 14, 14, 32)     │         36,864 │ conv4_block14_1_relu[… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block14_concat      │ (None, 14, 14, 704)    │              0 │ conv4_block13_concat[… │
│ (Concatenate)             │                        │                │ conv4_block14_2_conv[… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block15_0_bn        │ (None, 14, 14, 704)    │          2,816 │ conv4_block14_concat[… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block15_0_relu      │ (None, 14, 14, 704)    │              0 │ conv4_block15_0_bn[0]… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block15_1_conv      │ (None, 14, 14, 128)    │         90,112 │ conv4_block15_0_relu[… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block15_1_bn        │ (None, 14, 14, 128)    │            512 │ conv4_block15_1_conv[… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block15_1_relu      │ (None, 14, 14, 128)    │              0 │ conv4_block15_1_bn[0]… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block15_2_conv      │ (None, 14, 14, 32)     │         36,864 │ conv4_block15_1_relu[… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block15_concat      │ (None, 14, 14, 736)    │              0 │ conv4_block14_concat[… │
│ (Concatenate)             │                        │                │ conv4_block15_2_conv[… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block16_0_bn        │ (None, 14, 14, 736)    │          2,944 │ conv4_block15_concat[… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block16_0_relu      │ (None, 14, 14, 736)    │              0 │ conv4_block16_0_bn[0]… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block16_1_conv      │ (None, 14, 14, 128)    │         94,208 │ conv4_block16_0_relu[… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block16_1_bn        │ (None, 14, 14, 128)    │            512 │ conv4_block16_1_conv[… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block16_1_relu      │ (None, 14, 14, 128)    │              0 │ conv4_block16_1_bn[0]… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block16_2_conv      │ (None, 14, 14, 32)     │         36,864 │ conv4_block16_1_relu[… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block16_concat      │ (None, 14, 14, 768)    │              0 │ conv4_block15_concat[… │
│ (Concatenate)             │                        │                │ conv4_block16_2_conv[… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block17_0_bn        │ (None, 14, 14, 768)    │          3,072 │ conv4_block16_concat[… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block17_0_relu      │ (None, 14, 14, 768)    │              0 │ conv4_block17_0_bn[0]… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block17_1_conv      │ (None, 14, 14, 128)    │         98,304 │ conv4_block17_0_relu[… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block17_1_bn        │ (None, 14, 14, 128)    │            512 │ conv4_block17_1_conv[… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block17_1_relu      │ (None, 14, 14, 128)    │              0 │ conv4_block17_1_bn[0]… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block17_2_conv      │ (None, 14, 14, 32)     │         36,864 │ conv4_block17_1_relu[… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block17_concat      │ (None, 14, 14, 800)    │              0 │ conv4_block16_concat[… │
│ (Concatenate)             │                        │                │ conv4_block17_2_conv[… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block18_0_bn        │ (None, 14, 14, 800)    │          3,200 │ conv4_block17_concat[… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block18_0_relu      │ (None, 14, 14, 800)    │              0 │ conv4_block18_0_bn[0]… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block18_1_conv      │ (None, 14, 14, 128)    │        102,400 │ conv4_block18_0_relu[… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block18_1_bn        │ (None, 14, 14, 128)    │            512 │ conv4_block18_1_conv[… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block18_1_relu      │ (None, 14, 14, 128)    │              0 │ conv4_block18_1_bn[0]… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block18_2_conv      │ (None, 14, 14, 32)     │         36,864 │ conv4_block18_1_relu[… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block18_concat      │ (None, 14, 14, 832)    │              0 │ conv4_block17_concat[… │
│ (Concatenate)             │                        │                │ conv4_block18_2_conv[… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block19_0_bn        │ (None, 14, 14, 832)    │          3,328 │ conv4_block18_concat[… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block19_0_relu      │ (None, 14, 14, 832)    │              0 │ conv4_block19_0_bn[0]… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block19_1_conv      │ (None, 14, 14, 128)    │        106,496 │ conv4_block19_0_relu[… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block19_1_bn        │ (None, 14, 14, 128)    │            512 │ conv4_block19_1_conv[… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block19_1_relu      │ (None, 14, 14, 128)    │              0 │ conv4_block19_1_bn[0]… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block19_2_conv      │ (None, 14, 14, 32)     │         36,864 │ conv4_block19_1_relu[… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block19_concat      │ (None, 14, 14, 864)    │              0 │ conv4_block18_concat[… │
│ (Concatenate)             │                        │                │ conv4_block19_2_conv[… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block20_0_bn        │ (None, 14, 14, 864)    │          3,456 │ conv4_block19_concat[… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block20_0_relu      │ (None, 14, 14, 864)    │              0 │ conv4_block20_0_bn[0]… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block20_1_conv      │ (None, 14, 14, 128)    │        110,592 │ conv4_block20_0_relu[… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block20_1_bn        │ (None, 14, 14, 128)    │            512 │ conv4_block20_1_conv[… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block20_1_relu      │ (None, 14, 14, 128)    │              0 │ conv4_block20_1_bn[0]… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block20_2_conv      │ (None, 14, 14, 32)     │         36,864 │ conv4_block20_1_relu[… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block20_concat      │ (None, 14, 14, 896)    │              0 │ conv4_block19_concat[… │
│ (Concatenate)             │                        │                │ conv4_block20_2_conv[… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block21_0_bn        │ (None, 14, 14, 896)    │          3,584 │ conv4_block20_concat[… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block21_0_relu      │ (None, 14, 14, 896)    │              0 │ conv4_block21_0_bn[0]… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block21_1_conv      │ (None, 14, 14, 128)    │        114,688 │ conv4_block21_0_relu[… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block21_1_bn        │ (None, 14, 14, 128)    │            512 │ conv4_block21_1_conv[… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block21_1_relu      │ (None, 14, 14, 128)    │              0 │ conv4_block21_1_bn[0]… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block21_2_conv      │ (None, 14, 14, 32)     │         36,864 │ conv4_block21_1_relu[… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block21_concat      │ (None, 14, 14, 928)    │              0 │ conv4_block20_concat[… │
│ (Concatenate)             │                        │                │ conv4_block21_2_conv[… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block22_0_bn        │ (None, 14, 14, 928)    │          3,712 │ conv4_block21_concat[… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block22_0_relu      │ (None, 14, 14, 928)    │              0 │ conv4_block22_0_bn[0]… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block22_1_conv      │ (None, 14, 14, 128)    │        118,784 │ conv4_block22_0_relu[… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block22_1_bn        │ (None, 14, 14, 128)    │            512 │ conv4_block22_1_conv[… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block22_1_relu      │ (None, 14, 14, 128)    │              0 │ conv4_block22_1_bn[0]… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block22_2_conv      │ (None, 14, 14, 32)     │         36,864 │ conv4_block22_1_relu[… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block22_concat      │ (None, 14, 14, 960)    │              0 │ conv4_block21_concat[… │
│ (Concatenate)             │                        │                │ conv4_block22_2_conv[… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block23_0_bn        │ (None, 14, 14, 960)    │          3,840 │ conv4_block22_concat[… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block23_0_relu      │ (None, 14, 14, 960)    │              0 │ conv4_block23_0_bn[0]… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block23_1_conv      │ (None, 14, 14, 128)    │        122,880 │ conv4_block23_0_relu[… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block23_1_bn        │ (None, 14, 14, 128)    │            512 │ conv4_block23_1_conv[… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block23_1_relu      │ (None, 14, 14, 128)    │              0 │ conv4_block23_1_bn[0]… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block23_2_conv      │ (None, 14, 14, 32)     │         36,864 │ conv4_block23_1_relu[… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block23_concat      │ (None, 14, 14, 992)    │              0 │ conv4_block22_concat[… │
│ (Concatenate)             │                        │                │ conv4_block23_2_conv[… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block24_0_bn        │ (None, 14, 14, 992)    │          3,968 │ conv4_block23_concat[… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block24_0_relu      │ (None, 14, 14, 992)    │              0 │ conv4_block24_0_bn[0]… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block24_1_conv      │ (None, 14, 14, 128)    │        126,976 │ conv4_block24_0_relu[… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block24_1_bn        │ (None, 14, 14, 128)    │            512 │ conv4_block24_1_conv[… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block24_1_relu      │ (None, 14, 14, 128)    │              0 │ conv4_block24_1_bn[0]… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block24_2_conv      │ (None, 14, 14, 32)     │         36,864 │ conv4_block24_1_relu[… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv4_block24_concat      │ (None, 14, 14, 1024)   │              0 │ conv4_block23_concat[… │
│ (Concatenate)             │                        │                │ conv4_block24_2_conv[… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ pool4_bn                  │ (None, 14, 14, 1024)   │          4,096 │ conv4_block24_concat[… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ pool4_relu (Activation)   │ (None, 14, 14, 1024)   │              0 │ pool4_bn[0][0]         │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ pool4_conv (Conv2D)       │ (None, 14, 14, 512)    │        524,288 │ pool4_relu[0][0]       │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ pool4_pool                │ (None, 7, 7, 512)      │              0 │ pool4_conv[0][0]       │
│ (AveragePooling2D)        │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block1_0_bn         │ (None, 7, 7, 512)      │          2,048 │ pool4_pool[0][0]       │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block1_0_relu       │ (None, 7, 7, 512)      │              0 │ conv5_block1_0_bn[0][… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block1_1_conv       │ (None, 7, 7, 128)      │         65,536 │ conv5_block1_0_relu[0… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block1_1_bn         │ (None, 7, 7, 128)      │            512 │ conv5_block1_1_conv[0… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block1_1_relu       │ (None, 7, 7, 128)      │              0 │ conv5_block1_1_bn[0][… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block1_2_conv       │ (None, 7, 7, 32)       │         36,864 │ conv5_block1_1_relu[0… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block1_concat       │ (None, 7, 7, 544)      │              0 │ pool4_pool[0][0],      │
│ (Concatenate)             │                        │                │ conv5_block1_2_conv[0… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block2_0_bn         │ (None, 7, 7, 544)      │          2,176 │ conv5_block1_concat[0… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block2_0_relu       │ (None, 7, 7, 544)      │              0 │ conv5_block2_0_bn[0][… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block2_1_conv       │ (None, 7, 7, 128)      │         69,632 │ conv5_block2_0_relu[0… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block2_1_bn         │ (None, 7, 7, 128)      │            512 │ conv5_block2_1_conv[0… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block2_1_relu       │ (None, 7, 7, 128)      │              0 │ conv5_block2_1_bn[0][… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block2_2_conv       │ (None, 7, 7, 32)       │         36,864 │ conv5_block2_1_relu[0… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block2_concat       │ (None, 7, 7, 576)      │              0 │ conv5_block1_concat[0… │
│ (Concatenate)             │                        │                │ conv5_block2_2_conv[0… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block3_0_bn         │ (None, 7, 7, 576)      │          2,304 │ conv5_block2_concat[0… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block3_0_relu       │ (None, 7, 7, 576)      │              0 │ conv5_block3_0_bn[0][… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block3_1_conv       │ (None, 7, 7, 128)      │         73,728 │ conv5_block3_0_relu[0… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block3_1_bn         │ (None, 7, 7, 128)      │            512 │ conv5_block3_1_conv[0… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block3_1_relu       │ (None, 7, 7, 128)      │              0 │ conv5_block3_1_bn[0][… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block3_2_conv       │ (None, 7, 7, 32)       │         36,864 │ conv5_block3_1_relu[0… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block3_concat       │ (None, 7, 7, 608)      │              0 │ conv5_block2_concat[0… │
│ (Concatenate)             │                        │                │ conv5_block3_2_conv[0… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block4_0_bn         │ (None, 7, 7, 608)      │          2,432 │ conv5_block3_concat[0… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block4_0_relu       │ (None, 7, 7, 608)      │              0 │ conv5_block4_0_bn[0][… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block4_1_conv       │ (None, 7, 7, 128)      │         77,824 │ conv5_block4_0_relu[0… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block4_1_bn         │ (None, 7, 7, 128)      │            512 │ conv5_block4_1_conv[0… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block4_1_relu       │ (None, 7, 7, 128)      │              0 │ conv5_block4_1_bn[0][… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block4_2_conv       │ (None, 7, 7, 32)       │         36,864 │ conv5_block4_1_relu[0… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block4_concat       │ (None, 7, 7, 640)      │              0 │ conv5_block3_concat[0… │
│ (Concatenate)             │                        │                │ conv5_block4_2_conv[0… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block5_0_bn         │ (None, 7, 7, 640)      │          2,560 │ conv5_block4_concat[0… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block5_0_relu       │ (None, 7, 7, 640)      │              0 │ conv5_block5_0_bn[0][… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block5_1_conv       │ (None, 7, 7, 128)      │         81,920 │ conv5_block5_0_relu[0… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block5_1_bn         │ (None, 7, 7, 128)      │            512 │ conv5_block5_1_conv[0… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block5_1_relu       │ (None, 7, 7, 128)      │              0 │ conv5_block5_1_bn[0][… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block5_2_conv       │ (None, 7, 7, 32)       │         36,864 │ conv5_block5_1_relu[0… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block5_concat       │ (None, 7, 7, 672)      │              0 │ conv5_block4_concat[0… │
│ (Concatenate)             │                        │                │ conv5_block5_2_conv[0… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block6_0_bn         │ (None, 7, 7, 672)      │          2,688 │ conv5_block5_concat[0… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block6_0_relu       │ (None, 7, 7, 672)      │              0 │ conv5_block6_0_bn[0][… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block6_1_conv       │ (None, 7, 7, 128)      │         86,016 │ conv5_block6_0_relu[0… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block6_1_bn         │ (None, 7, 7, 128)      │            512 │ conv5_block6_1_conv[0… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block6_1_relu       │ (None, 7, 7, 128)      │              0 │ conv5_block6_1_bn[0][… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block6_2_conv       │ (None, 7, 7, 32)       │         36,864 │ conv5_block6_1_relu[0… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block6_concat       │ (None, 7, 7, 704)      │              0 │ conv5_block5_concat[0… │
│ (Concatenate)             │                        │                │ conv5_block6_2_conv[0… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block7_0_bn         │ (None, 7, 7, 704)      │          2,816 │ conv5_block6_concat[0… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block7_0_relu       │ (None, 7, 7, 704)      │              0 │ conv5_block7_0_bn[0][… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block7_1_conv       │ (None, 7, 7, 128)      │         90,112 │ conv5_block7_0_relu[0… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block7_1_bn         │ (None, 7, 7, 128)      │            512 │ conv5_block7_1_conv[0… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block7_1_relu       │ (None, 7, 7, 128)      │              0 │ conv5_block7_1_bn[0][… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block7_2_conv       │ (None, 7, 7, 32)       │         36,864 │ conv5_block7_1_relu[0… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block7_concat       │ (None, 7, 7, 736)      │              0 │ conv5_block6_concat[0… │
│ (Concatenate)             │                        │                │ conv5_block7_2_conv[0… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block8_0_bn         │ (None, 7, 7, 736)      │          2,944 │ conv5_block7_concat[0… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block8_0_relu       │ (None, 7, 7, 736)      │              0 │ conv5_block8_0_bn[0][… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block8_1_conv       │ (None, 7, 7, 128)      │         94,208 │ conv5_block8_0_relu[0… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block8_1_bn         │ (None, 7, 7, 128)      │            512 │ conv5_block8_1_conv[0… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block8_1_relu       │ (None, 7, 7, 128)      │              0 │ conv5_block8_1_bn[0][… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block8_2_conv       │ (None, 7, 7, 32)       │         36,864 │ conv5_block8_1_relu[0… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block8_concat       │ (None, 7, 7, 768)      │              0 │ conv5_block7_concat[0… │
│ (Concatenate)             │                        │                │ conv5_block8_2_conv[0… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block9_0_bn         │ (None, 7, 7, 768)      │          3,072 │ conv5_block8_concat[0… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block9_0_relu       │ (None, 7, 7, 768)      │              0 │ conv5_block9_0_bn[0][… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block9_1_conv       │ (None, 7, 7, 128)      │         98,304 │ conv5_block9_0_relu[0… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block9_1_bn         │ (None, 7, 7, 128)      │            512 │ conv5_block9_1_conv[0… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block9_1_relu       │ (None, 7, 7, 128)      │              0 │ conv5_block9_1_bn[0][… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block9_2_conv       │ (None, 7, 7, 32)       │         36,864 │ conv5_block9_1_relu[0… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block9_concat       │ (None, 7, 7, 800)      │              0 │ conv5_block8_concat[0… │
│ (Concatenate)             │                        │                │ conv5_block9_2_conv[0… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block10_0_bn        │ (None, 7, 7, 800)      │          3,200 │ conv5_block9_concat[0… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block10_0_relu      │ (None, 7, 7, 800)      │              0 │ conv5_block10_0_bn[0]… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block10_1_conv      │ (None, 7, 7, 128)      │        102,400 │ conv5_block10_0_relu[… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block10_1_bn        │ (None, 7, 7, 128)      │            512 │ conv5_block10_1_conv[… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block10_1_relu      │ (None, 7, 7, 128)      │              0 │ conv5_block10_1_bn[0]… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block10_2_conv      │ (None, 7, 7, 32)       │         36,864 │ conv5_block10_1_relu[… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block10_concat      │ (None, 7, 7, 832)      │              0 │ conv5_block9_concat[0… │
│ (Concatenate)             │                        │                │ conv5_block10_2_conv[… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block11_0_bn        │ (None, 7, 7, 832)      │          3,328 │ conv5_block10_concat[… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block11_0_relu      │ (None, 7, 7, 832)      │              0 │ conv5_block11_0_bn[0]… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block11_1_conv      │ (None, 7, 7, 128)      │        106,496 │ conv5_block11_0_relu[… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block11_1_bn        │ (None, 7, 7, 128)      │            512 │ conv5_block11_1_conv[… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block11_1_relu      │ (None, 7, 7, 128)      │              0 │ conv5_block11_1_bn[0]… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block11_2_conv      │ (None, 7, 7, 32)       │         36,864 │ conv5_block11_1_relu[… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block11_concat      │ (None, 7, 7, 864)      │              0 │ conv5_block10_concat[… │
│ (Concatenate)             │                        │                │ conv5_block11_2_conv[… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block12_0_bn        │ (None, 7, 7, 864)      │          3,456 │ conv5_block11_concat[… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block12_0_relu      │ (None, 7, 7, 864)      │              0 │ conv5_block12_0_bn[0]… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block12_1_conv      │ (None, 7, 7, 128)      │        110,592 │ conv5_block12_0_relu[… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block12_1_bn        │ (None, 7, 7, 128)      │            512 │ conv5_block12_1_conv[… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block12_1_relu      │ (None, 7, 7, 128)      │              0 │ conv5_block12_1_bn[0]… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block12_2_conv      │ (None, 7, 7, 32)       │         36,864 │ conv5_block12_1_relu[… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block12_concat      │ (None, 7, 7, 896)      │              0 │ conv5_block11_concat[… │
│ (Concatenate)             │                        │                │ conv5_block12_2_conv[… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block13_0_bn        │ (None, 7, 7, 896)      │          3,584 │ conv5_block12_concat[… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block13_0_relu      │ (None, 7, 7, 896)      │              0 │ conv5_block13_0_bn[0]… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block13_1_conv      │ (None, 7, 7, 128)      │        114,688 │ conv5_block13_0_relu[… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block13_1_bn        │ (None, 7, 7, 128)      │            512 │ conv5_block13_1_conv[… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block13_1_relu      │ (None, 7, 7, 128)      │              0 │ conv5_block13_1_bn[0]… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block13_2_conv      │ (None, 7, 7, 32)       │         36,864 │ conv5_block13_1_relu[… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block13_concat      │ (None, 7, 7, 928)      │              0 │ conv5_block12_concat[… │
│ (Concatenate)             │                        │                │ conv5_block13_2_conv[… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block14_0_bn        │ (None, 7, 7, 928)      │          3,712 │ conv5_block13_concat[… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block14_0_relu      │ (None, 7, 7, 928)      │              0 │ conv5_block14_0_bn[0]… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block14_1_conv      │ (None, 7, 7, 128)      │        118,784 │ conv5_block14_0_relu[… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block14_1_bn        │ (None, 7, 7, 128)      │            512 │ conv5_block14_1_conv[… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block14_1_relu      │ (None, 7, 7, 128)      │              0 │ conv5_block14_1_bn[0]… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block14_2_conv      │ (None, 7, 7, 32)       │         36,864 │ conv5_block14_1_relu[… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block14_concat      │ (None, 7, 7, 960)      │              0 │ conv5_block13_concat[… │
│ (Concatenate)             │                        │                │ conv5_block14_2_conv[… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block15_0_bn        │ (None, 7, 7, 960)      │          3,840 │ conv5_block14_concat[… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block15_0_relu      │ (None, 7, 7, 960)      │              0 │ conv5_block15_0_bn[0]… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block15_1_conv      │ (None, 7, 7, 128)      │        122,880 │ conv5_block15_0_relu[… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block15_1_bn        │ (None, 7, 7, 128)      │            512 │ conv5_block15_1_conv[… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block15_1_relu      │ (None, 7, 7, 128)      │              0 │ conv5_block15_1_bn[0]… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block15_2_conv      │ (None, 7, 7, 32)       │         36,864 │ conv5_block15_1_relu[… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block15_concat      │ (None, 7, 7, 992)      │              0 │ conv5_block14_concat[… │
│ (Concatenate)             │                        │                │ conv5_block15_2_conv[… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block16_0_bn        │ (None, 7, 7, 992)      │          3,968 │ conv5_block15_concat[… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block16_0_relu      │ (None, 7, 7, 992)      │              0 │ conv5_block16_0_bn[0]… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block16_1_conv      │ (None, 7, 7, 128)      │        126,976 │ conv5_block16_0_relu[… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block16_1_bn        │ (None, 7, 7, 128)      │            512 │ conv5_block16_1_conv[… │
│ (BatchNormalization)      │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block16_1_relu      │ (None, 7, 7, 128)      │              0 │ conv5_block16_1_bn[0]… │
│ (Activation)              │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block16_2_conv      │ (None, 7, 7, 32)       │         36,864 │ conv5_block16_1_relu[… │
│ (Conv2D)                  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ conv5_block16_concat      │ (None, 7, 7, 1024)     │              0 │ conv5_block15_concat[… │
│ (Concatenate)             │                        │                │ conv5_block16_2_conv[… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ bn (BatchNormalization)   │ (None, 7, 7, 1024)     │          4,096 │ conv5_block16_concat[… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ relu (Activation)         │ (None, 7, 7, 1024)     │              0 │ bn[0][0]               │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ global_average_pooling2d… │ (None, 1024)           │              0 │ relu[0][0]             │
│ (GlobalAveragePooling2D)  │                        │                │                        │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ dense_16 (Dense)          │ (None, 512)            │        524,800 │ global_average_poolin… │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ dropout_8 (Dropout)       │ (None, 512)            │              0 │ dense_16[0][0]         │
├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤
│ dense_17 (Dense)          │ (None, 1)              │            513 │ dropout_8[0][0]        │
└───────────────────────────┴────────────────────────┴────────────────┴────────────────────────┘

 Total params: 7,562,817 (28.85 MB)

 Trainable params: 525,313 (2.00 MB)

 Non-trainable params: 7,037,504 (26.85 MB)

# Early stopping
early_stopping = EarlyStopping(monitor='val_loss', patience=5, restore_best_weights=True)

# Start timer
start_time = time.time()

# Train the model
history = densenet_model.fit(
    train_generator,
    epochs=20,
    validation_data=test_generator,
    callbacks=[early_stopping]
)

# End timer
end_time = time.time()

# Calculate training duration
training_time = end_time - start_time
minutes, seconds = divmod(training_time, 60)
print("-" * 120)
print(f"Training completed in {int(minutes)} minutes and {int(seconds)} seconds.")

# Plot accuracy and loss graphs
plt.figure(figsize=(12, 3))

# Plot Accuracy
plt.subplot(1, 2, 1)
plt.plot(history.history['accuracy'], label='Train Accuracy')
plt.plot(history.history['val_accuracy'], label='Test Accuracy')
plt.xlabel('Epochs')
plt.ylabel('Accuracy')
plt.legend()
plt.title('\nTraining vs Test Accuracy\n')

# Plot Loss
plt.subplot(1, 2, 2)
plt.plot(history.history['loss'], label='Train Loss')
plt.plot(history.history['val_loss'], label='Test Loss')
plt.xlabel('Epochs')
plt.ylabel('Loss')
plt.legend()
plt.title('\nTraining vs Test Loss\n')

plt.show()

Epoch 1/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 46s 165ms/step - accuracy: 0.7200 - loss: 0.5725 - val_accuracy: 0.8530 - val_loss: 0.3185
Epoch 2/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 11s 66ms/step - accuracy: 0.8330 - loss: 0.3597 - val_accuracy: 0.8959 - val_loss: 0.2791
Epoch 3/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 11s 65ms/step - accuracy: 0.8529 - loss: 0.3314 - val_accuracy: 0.8790 - val_loss: 0.2656
Epoch 4/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 12s 68ms/step - accuracy: 0.8747 - loss: 0.2981 - val_accuracy: 0.8239 - val_loss: 0.3500
Epoch 5/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 11s 66ms/step - accuracy: 0.8802 - loss: 0.2800 - val_accuracy: 0.9096 - val_loss: 0.2137
Epoch 6/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 12s 68ms/step - accuracy: 0.8874 - loss: 0.2529 - val_accuracy: 0.9219 - val_loss: 0.1939
Epoch 7/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 11s 67ms/step - accuracy: 0.9138 - loss: 0.2070 - val_accuracy: 0.9081 - val_loss: 0.2228
Epoch 8/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 11s 66ms/step - accuracy: 0.9064 - loss: 0.2193 - val_accuracy: 0.9326 - val_loss: 0.1715
Epoch 9/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 11s 66ms/step - accuracy: 0.9262 - loss: 0.1762 - val_accuracy: 0.9372 - val_loss: 0.1665
Epoch 10/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 12s 69ms/step - accuracy: 0.9205 - loss: 0.1869 - val_accuracy: 0.9326 - val_loss: 0.1491
Epoch 11/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 11s 68ms/step - accuracy: 0.9273 - loss: 0.1787 - val_accuracy: 0.9433 - val_loss: 0.1388
Epoch 12/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 11s 67ms/step - accuracy: 0.9388 - loss: 0.1529 - val_accuracy: 0.9495 - val_loss: 0.1383
Epoch 13/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 11s 66ms/step - accuracy: 0.9455 - loss: 0.1378 - val_accuracy: 0.9372 - val_loss: 0.1607
Epoch 14/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 11s 67ms/step - accuracy: 0.9418 - loss: 0.1455 - val_accuracy: 0.9510 - val_loss: 0.1249
Epoch 15/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 11s 67ms/step - accuracy: 0.9429 - loss: 0.1541 - val_accuracy: 0.9510 - val_loss: 0.1188
Epoch 16/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 11s 67ms/step - accuracy: 0.9511 - loss: 0.1229 - val_accuracy: 0.9632 - val_loss: 0.1085
Epoch 17/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 11s 66ms/step - accuracy: 0.9599 - loss: 0.1091 - val_accuracy: 0.9617 - val_loss: 0.0980
Epoch 18/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 11s 66ms/step - accuracy: 0.9600 - loss: 0.1032 - val_accuracy: 0.9188 - val_loss: 0.1712
Epoch 19/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 11s 66ms/step - accuracy: 0.9445 - loss: 0.1299 - val_accuracy: 0.9587 - val_loss: 0.0994
Epoch 20/20
164/164 ━━━━━━━━━━━━━━━━━━━━ 12s 69ms/step - accuracy: 0.9637 - loss: 0.1067 - val_accuracy: 0.9694 - val_loss: 0.0869
------------------------------------------------------------------------------------------------------------------------
Training completed in 4 minutes and 22 seconds.

# Predict the test set
y_pred_prob = densenet_model.predict(test_generator)
y_pred = (y_pred_prob > 0.5).astype(int)

# Get true labels
y_true = test_generator.classes

# Confusion Matrix
conf_matrix = confusion_matrix(y_true, y_pred)
plt.figure(figsize=(4, 3))
sns.heatmap(conf_matrix, annot=True, fmt='d', cmap='Blues', 
            xticklabels=['Healthy', 'Parkinson'], 
            yticklabels=['Healthy', 'Parkinson'], 
            cbar=False)
plt.xlabel('Predicted')
plt.ylabel('Actual')
plt.title('\nTest Data\n')
plt.show()

# Evaluation Metrics
accuracy = accuracy_score(y_true, y_pred)
f1 = f1_score(y_true, y_pred)
auc = roc_auc_score(y_true, y_pred_prob)

print(f'Accuracy: {accuracy:.2f}')
print(f'F1 Score: {f1:.2f}')
print(f'AUC Score: {auc:.2f}')

41/41 ━━━━━━━━━━━━━━━━━━━━ 16s 222ms/step

Accuracy: 0.97
F1 Score: 0.97
AUC Score: 1.00

---

Step 2: Traditional Models

1. Support Vector Machine: Uses the RBF kernel (gamma='scale') with C=1.0 to find the optimal hyperplane for separating classes, effective for non-linear decision boundaries.

2. K-Nearest Neighbor: Classifies samples based on the majority class of their 5 nearest neighbors (n_neighbors=5), using distance metrics for similarity.

3. Random Forest: An ensemble method with 100 decision trees (n_estimators=100) to improve classification accuracy and reduce overfitting through randomness.

from sklearn.svm import SVC
from sklearn.neighbors import KNeighborsClassifier
from sklearn.ensemble import RandomForestClassifier

# Define the models dictionary
models = {
    "SVM": SVC(kernel='rbf', C=1.0, gamma='scale'),
    "KNN": KNeighborsClassifier(n_neighbors=5),
    "Random Forest": RandomForestClassifier(n_estimators=100, random_state=42)
}

# Loop through each model, train and evaluate
for name, model in models.items():
    print("-" * 25)
    print(f"Model: {name}")
    
    # Train the model
    model.fit(X_train, y_train)
    
    # Predict on the test set
    y_pred = model.predict(X_test)
    
    # Evaluate accuracy
    accuracy = accuracy_score(y_test, y_pred)
    print(f'Accuracy: {accuracy:.2f}')
    
    # Compute confusion matrix
    conf_matrix = confusion_matrix(y_test, y_pred)
    print(f"Confusion Matrix:")
    print(conf_matrix)

-------------------------
Model: SVM
Accuracy: 0.92
Confusion Matrix:
[[317  10]
 [ 41 285]]
-------------------------
Model: KNN
Accuracy: 0.86
Confusion Matrix:
[[321   6]
 [ 84 242]]
-------------------------
Model: Random Forest
Accuracy: 0.88
Confusion Matrix:
[[313  14]
 [ 65 261]]

---

5 Results and Analysis

1. Performance Comparison:

• Best Performing Model: MobileNetV2:

  • MobileNetV2 had the best overall performance (0.97) with a balanced confusion matrix, showing high sensitivity and specificity.
  • DenseNet121 also performed exceptionally well (0.97), with slightly more misclassified samples compared to MobileNetV2.
    
    

• Performance Overview:

  • CNN performed well (0.96), slightly behind MobileNetV2 and DenseNet121.
  • VGG16 had lower accuracy (0.92) due to more misclassified Parkinson cases.
  • Traditional models (SVM, KNN, RF) performed significantly worse, with KNN having the lowest accuracy (0.86), suggesting that deep learning models are better suited for this task.
    
    

• Conclusion: Deep learning models, especially MobileNetV2 and DenseNet121, significantly outperformed traditional ML models, demonstrating their effectiveness in Parkinson’s disease detection using hand-drawn images.

Model	Accuracy	F1-Score	AUC Score	Confusion Matrix
CNN	0.96	0.96	0.98	[318 09] [15 311]
MobileNet	0.97	0.97	1.00	[317 10] [08 318]
VGG	0.92	0.91	0.98	[323 04] [51 275]
DenseNet	0.97	0.97	1.00	[325 02] [18 308]
SVM	0.92	0.92	0.92	[317 10] [41 285]
KNN	0.86	0.84	0.86	[321 06] [84 242]
RF	0.88	0.87	0.88	[313 14] [65 261]

# Model names
models = ["CNN", "DenseNet", "MobileNet", "SVM", "VGG", "RF", "KNN"]

# Calculate error rates
total_samples = 653
error_samples = {"CNN": 24, "MobileNet": 18, "VGG": 55, "DenseNet": 20, "SVM": 51, "KNN": 90, "RF": 79}
error_rate = [round((error_samples[model] / total_samples) * 100, 2) for model in models]

# Sort models by error rates in ascending order
sorted_indices = np.argsort(error_rate)
models = [models[i] for i in sorted_indices]
error_rate = [error_rate[i] for i in sorted_indices]

# Plot bar chart
plt.figure(figsize=(7, 3))
plt.bar(models, error_rate, color='lightsalmon', edgecolor='black', width=0.4)
plt.title("\nError Rate of Models\n", fontsize=12)
plt.ylim(0, max(error_rate) + 5)  # Set y-axis range
plt.yticks(ticks=range(0, int(max(error_rate)) + 10, 5), labels=[f"{tick}%" for tick in range(0, int(max(error_rate)) + 10, 5)])

# Display values on each bar
for i, error in enumerate(error_rate):
    plt.text(i, error + 0.5, f"{error}%", ha='center', fontsize=10)

plt.grid(axis='y', linestyle='--', alpha=0.5)
# plt.savefig('error_rate.png', dpi=300)
plt.show()

---

2. Error Analysis:

• Misclassified Images (using MobileNetV2 as a sample):

  • Based on the classification results, there are 18 misclassified samples:

    • (True: 0, Pred: 1) - False Positives (FP): 10 samples (5 spirals, 5 waves)
    • (True: 1, Pred: 0) - False Negatives (FN): 8 samples (6 spirals, 2 waves)
      
      

  • This suggests that spirals tend to be more challenging for the model to classify correctly.

• Possible Reasons for Misclassification:

  • False Positives (FP):

    • Natural variations in handwriting styles: Some healthy individuals may naturally have slower or less steady hand movements, creating drawings that resemble those of PD patients.
    • Noise in the dataset: Differences in image quality, pen pressure, lighting conditions, or stroke thickness could lead to misinterpretations by the model.
    • Aging effects: Elderly individuals without Parkinson’s may exhibit slight motor impairments, causing their drawings to resemble those of PD patients.
      
      

  • False Negatives (FN):

    • Early-stage PD symptoms: Patients in early stages of Parkinson’s may still retain relatively steady hand control, making their drawings appear normal.
    • Compensatory effort: Patients might actively try to draw more slowly and carefully, reducing tremor effects and making their drawings appear more like those of healthy individuals.
    • Symptom variability: Motor symptoms fluctuate, meaning that some PD patients may have better motor control on the day of testing, leading to less distinguishable patterns.

# Identify misclassified samples
misclassified_indices = [i for i, (true, pred) in enumerate(zip(y_test, y_pred)) if true != pred]
print(f"Number of misclassified samples: {len(misclassified_indices)}")

# Separate them into two categories
misclassified_0_as_1 = [idx for idx in misclassified_indices if y_test[idx] == 0 and y_pred[idx] == 1]
misclassified_1_as_0 = [idx for idx in misclassified_indices if y_test[idx] == 1 and y_pred[idx] == 0]

# Print the results
print(f"Misclassified samples (True: 0, Pred: 1): {len(misclassified_0_as_1)}, Indices: {misclassified_0_as_1}")
print(f"Misclassified samples (True: 1, Pred: 0): {len(misclassified_1_as_0)}, Indices: {misclassified_1_as_0}")

Number of misclassified samples: 18
Misclassified samples (True: 0, Pred: 1): 10, Indices: [0, 32, 60, 83, 89, 194, 219, 247, 265, 282]
Misclassified samples (True: 1, Pred: 0): 8, Indices: [404, 522, 525, 559, 585, 598, 603, 605]

# Visualize misclassified images
def visualize_images_grid(misclassified_0_as_1, misclassified_1_as_0, test_images):
    num_cols_0_as_1 = 10  # Number of images in the first row
    num_cols_1_as_0 = 8   # Number of images in the second row

    plt.figure(figsize=(20, 8))  # Adjust figure size

    # Plot the first row: True: 0, Pred: 1 (False Positive)
    for i, idx in enumerate(misclassified_0_as_1):
        img_path = test_images[idx]
        img = Image.open(img_path)
        plt.subplot(2, num_cols_0_as_1, i + 1)
        plt.imshow(img)
        plt.axis('off')
        plt.title(f"Idx: {idx} (FP)", fontsize=10)

    # Plot the second row: True: 1, Pred: 0 (False Negative)
    for i, idx in enumerate(misclassified_1_as_0):
        img_path = test_images[idx]
        img = Image.open(img_path)
        plt.subplot(2, num_cols_0_as_1, num_cols_0_as_1 + i + 1)
        plt.imshow(img)
        plt.axis('off')
        plt.title(f"Idx: {idx} (FN)", fontsize=10)

    plt.tight_layout()
    plt.suptitle("\nMisclassified Images\n", fontsize=16)
    # plt.savefig('misclassified_images.png', dpi=300)
    plt.show()

visualize_images_grid(misclassified_0_as_1, misclassified_1_as_0, test_images)

---

6 Discussion and Conclusion

Discussion

1. Model Strengths and Weaknesses:

• Deep Learning:

  • Strengths:

    • Automatically learns hierarchical features from data.
    • Achieves high accuracy in complex image-based tasks.
    • Highly scalable, making it suitable for large-scale applications.
      
      

  • Weaknesses:

    • High computational cost and memory usage, often requiring GPUs.
    • Requires large datasets for optimal performance.
    • Longer training times compared to traditional models.
      
      

  • MobileNetV2 achieves the lowest error rate (2.76%) with the smallest model size (9.87MB), making it ideal for lightweight applications. VGG16, despite its strong feature extraction capabilities, has the highest error rate (8.42%), suggesting inefficiencies in this task. CNN and DenseNet121 balance accuracy and resource usage but require more memory than MobileNetV2.
    
    

Model	Total Parameters	Model Size (MB)	Training Time	Error Rate (%)
CNN	22,244,929	84.86	2m 59s	3.68
MobileNet	2,586,177	9.87	3m 39s	2.76
VGG	14,977,857	57.14	4m 47s	8.42
DenseNet	7,562,817	28.85	4m 22s	3.06

• Traditional Models:

  • Strengths:

    • Perform well on smaller datasets without the need for extensive computational resources.
    • Easier to train and tune compared to deep learning models.
    • Provide interpretable decision-making processes in certain cases.
      
      

  • Weaknesses:

    • Struggle to capture complex patterns in high-dimensional data.
    • Less robust to variations compared to deep learning models.
    • Require manual feature extraction, which can be time-consuming
      
      

  • Traditional machine learning models, such as SVM, KNN, and RF, can remain competitive in scenarios with limited resources or smaller datasets. SVM, for example, achieves a lower error rate (7.81%) compared to VGG16 (8.42%), likely because it is more effective in limited data settings and focuses on finding optimal decision boundaries. Their simplicity and lower risk of overfitting make traditional models advantageous over deep learning in such cases.

---

2. Limitations and Future Work:

• Limitations:

  • Result Limitations:

    • Model Performance in Specific Scenarios: While most models performed well, VGG16 showed inferior performance compared to other CNN models, likely due to its large size and architectural inefficiencies for this task. Additionally, class imbalances or biased data distributions may still impact model performance.

    • Training Time and Resource Demands: Deeper models like VGG1 and DenseNet121 required significantly more computational resources and longer training times, making them less efficient for deployment on resource-constrained systems.
      
      

  • Methodological Limitations:

    • Model and Dataset Constraints: While data augmentation was applied, the dataset size (3,264 images) remains a constraint, potentially limiting model generalization, especially for edge cases. Collecting a larger and more diverse dataset could further enhance robustness.

    • Potential for Alternative Architectures: Exploring alternative architectures, such as lighter CNN variants or hybrid models, may improve efficiency and accuracy while reducing computational costs.
      
      

• Future Work:

  • Data:

    • Introduce a larger dataset, particularly focusing on underrepresented hand-drawn patterns.
    • Consider incorporating multimodal data, such as handwriting dynamics or drawing speed, to improve classification performance.
      
      

  • Model:

    • Explore more efficient models, such as lightweight deep learning architectures, to optimize performance and reduce computational demands.
    • Experiment with alternative loss functions or regularization techniques to enhance classification accuracy and robustness.
    • Integrate time-series analysis or use deeper network structures to better capture subtle motor impairments.
      
      

  • Application:

    • Investigate deployment strategies for clinical environments, ensuring the model can be tested and validated under real-world conditions.

---

Conclusion

In this project, I evaluated CNN, MobileNet, VGG, and DenseNet for classifying hand-drawn patterns as indicators of PD or non-PD. MobileNet performed best with 2.76% error rate and a compact model size, making it the most efficient. To compare with deep learning, I applied SVM, KNN, and Random Forest as baselines. SVM outperformed VGG, showing that traditional models remain competitive in small-data scenarios. Overall, this study highlights the potential of deep learning in analyzing motor impairments through hand-drawn patterns and provides insights into model selection for Parkinson’s disease classification.

GitHub Repository Link

https://github.com/d93xup60126/Deep_Learning_PD_Detection

---